Essentials of Organic Chemistry

86 STEREOCHEMISTRY

We have seen earlier that enantiomers are chem- is a diastereoisomer that differs in chirality at only one
ically identical except in optical properties, although centre. Thus, (−)-pseudoephedrine is the 2-epimer of
biological properties may be different (see Box 3.7). (−)-ephedrine, and (+)-pseudoephedrine is the 1-epimer
On the other hand, diastereoisomers have different of (−)-ephedrine.
physical and chemical properties, and probably dif-
ferent biological properties as well. As a result, they The epimer terminology is of greater value when
are considered a completely different chemical entity, there are more than two chiral centres in the molecule.
and are often given a different chemical name. The Suppose we have a compound with three chiral centres,
(1S,2S) and (1R,2R) isomers are thus known as (+)- at positions 2, 3, and 4 in some unspecified carbon
pseudoephedrine and (−)-pseudoephedrine respectively. chain, with configurations 2R,3R,4S. There would thus
Interestingly, (+)-pseudoephedrine has similar biologi- exist a total of 23 = 8 configurational isomers. The
cal properties to (−)-ephedrine, and it is used as a bron- enantiomer would have the configuration 2S,3S,4R, i.e.
chodilator and decongestant drug in the same way as changing the configuration at all centres. The 2S,3R,4S
ephedrine. diastereoisomer we could then refer to as ‘the 2-epimer’,
and the 2R,3S,4S diastereoisomer as ‘the 3-epimer’,
One more useful piece of terminology can be since we have changed the stereochemistry at just one
introduced here. This is the term epimer. An epimer centre, keeping other configurations the same.

Box 3.10

Drawing enantiomers and epimers: 6-aminopenicillanic acid

The structure of the natural isomer of 6-aminopenicillanic acid is shown. You are asked to draw the structure
of its enantiomer and its 6-epimer.

mirror

H2N H H NH2 H2N H
O
S S ≡ S it is easier to change
15 6 51 wedges/dots than to draw the
6 51 2 6 OO 2 mirror image; reverse the
2 3 7 3 configuration at all centres
7 N4 3 4N 7 N4

CO2H HO2C CO2H

(3S,5R,6R)- enantiomer
6-aminopenicillanic acid
(3R,5S,6S)-
6-aminopenicillanic acid

H2N H

S
6 51
2 change the configuration
7 N4 3 at one centre only, by
reversing wedge/dots
O
CO2H

6-epimer

(3S,5R,6S)-
6-aminopenicillanic acid

The enantiomer will have the configuration changed at all chiral centres, whereas the 6-epimer retains all
configurations except for that at position 6. Note that it is not necessary to draw the mirror image compound for
the enantiomer, just reverse the wedge–dot relationship for the bonds at each chiral centre. This is much easier
and less prone to errors whilst transcribing the structure.

CONFIGURATIONAL ISOMERS 87

Now for a rather important point. In a compound 2-methylcyclopropanecarboxylic acid
such as (−)-ephedrine there are going to be many
different conformations as a result of rotation about mirror
the central C–C bond; three of them are shown here,
the energetically most favourable staggered conformer H3C R H H S CH3
with all large groups anti, a less favourable staggered R S
conformer, and a high-energy eclipsed version. 2 2
1 1
H H
CO2H HO2C

HO HO H (+)- and (–)-trans enantiomers
H H
CH3 mirror
1 2 CH3
1 2 HNHCH3 Ph NHCH3 H3C R S CO2H HO2C R S CH3

Ph 1R,2S 2 1 1 2
(–)-ephedrine
1R,2S H H H H
(–)-ephedrine
(+)- and (–)-cis enantiomers

favourable less favourable However, in a cyclohexane system we also need to
staggered conformer: staggered conformer consider the conformational mobility that generates two
large groups all anti different chair forms of the ring (see Section 3.3.2). Let
us consider 3-methylcyclohexanecarboxylic acid. This
HO NHHCH3 has two chiral centres, and thus there are four configu-
H rational stereoisomers. These are the enantiomeric forms
of the trans and cis isomers.

a change in conformation does 12 3-methylcyclohexanecarboxylic acid
not affect configuration
Ph CH3

1R,2S 31 trans
(–)-ephedrine

unfavourable CO2H
eclipsed conformer

However, note carefully that changing the confor- CH3 mirror CH3
mation does not affect the spatial sequence about the CO2H HO2C
chiral centres, i.e. it does not change the configura-
tion at either chiral centre. This seems a trivial and H3C CH3
rather obvious statement, and indeed it probably is
in the case of acyclic compounds. It is when we CO2H HO2C
move on to cyclic compounds that we need to remem-
ber this fundamental concept, because a common mis- (+)- and (–)-trans enantiomers; two chair
take is to confuse conformation and configuration (see conformations are shown for each, the favoured one
Box 3.11). is likely to have the larger carboxylic acid group
equatorial − note that the mirror image relationship
The same stereochemical principles are going to apply is readily apparent in both conformers
to both acyclic and cyclic compounds. With simple
cyclic compounds that have little or no conformational Care: this shows two interconvertible conformers for
mobility, it is easier to follow what is going on. each of the two non-interconvertible enantiomers
Consider a disubstituted cyclopropane system. As in the
acyclic examples, there are four different configurational
stereoisomers possible, comprising two pairs of
enantiomers. No conformational mobility is possible
here.

88 STEREOCHEMISTRY

3-methylcyclohexanecarboxylic acid 4-methylcyclohexanecarboxylic acid

31 cis 4

CO2H 1 trans
CO2H

mirror

CH3 CO2H mirror
HO2C
CH3 CO2H ≡ HO2C
H3C CH3
CH3 CH3

≡

H3C CO2H HO2C CH3 CO2H HO2C

(+)- and (–)-cis enantiomers; two chair CO2H plane of symmetry
conformations are shown for each, the favoured
diequatorial and the unfavoured diaxial − note that 4
the mirror image relationship is readily apparent in
both conformers 1 cis

Care: this shows two interconvertible conformers for CO2H
each of the two non-interconvertible enantiomers
mirror
Each isomer can also adopt a different chair confor-
mation as a consequence of ring flip (see Section 3.3.2). CO2H HO2C
We thus can write down eight possible stereoisomers,
comprised of two interconvertible conformers for each H3C ≡ CH3
of the four non-interconvertible configurational iso- CH3 CH3
mers. Put another way, there are four configurational
isomers (22 = 4), but each can exist as two possi- ≡
ble conformational isomers. Note that you can also CO2H HO2C
see the mirror image relationship in the conforma-
tional isomers. Of course, in practice, some conform- CO2H plane of symmetry
ers are not going to be energetically favourable. The
cis compound has favoured diequatorial and unfavoured exist; these are geometric isomers (see Section 3.4.3)
diaxial conformers. The trans compound has one equa- and can still be regarded as diastereoisomers.
torial and one axial substituent; we can assume that
the larger carboxylic acid group will prefer to be We can spot this type of situation by looking for
equatorial. symmetry in the molecule. Both cis- and trans-4-
methylcyclohexanecarboxylic acid isomers have a plane
Do appreciate that cyclohexane rings with 1,2- or of symmetry, and, as we saw for simple tetrahedral
1,3-substitution fit into the above discussions; however, carbons (see Section 3.4.1), this symmetry means the
if we have 1,4-substitution there are no chiral centres molecule is achiral.
in the molecule, since two of the groups are the same
at each possible site! However, cis and trans forms still

CONFIGURATIONAL ISOMERS 89

Box 3.11

Configurations and conformations: avoiding confusion

At this stage, a word of caution: do not confuse conformation with configuration. Different conformations
interconvert easily; different configurations do not interconvert without some bond-breaking process. We
commented above that changing the conformation did not affect the spatial sequence about chiral centres, and used
ephedrine as a rather trivial and obvious example. Rotation about single bonds did not change the configuration
at either chiral centre.

To emphasize this point, look at the following relationships for trans-3-methylcyclohexyl bromide.

don't confuse conformation with configuration H eq
CH3
ax Br
H CH3 ax Br
Br Br H ring flip
eq H

these do not conformer:
interconvert has same configuration
at each centre
ax

Br H H

H CH3 ring flip CH3
eq ax
Br Br Br
eq
enantiomer:
has different configuration H
at each centre

Ring flip of the upper left structure produces an alternative conformer. Ring flip does not change the
configuration. The axial–equatorial relationship (conformation) is modified, but the up–down relationship
(configuration) is still there. The enantiomer of this structure has the alternative configuration at both chiral
centres, but it cannot be produced from the first structure by any simple isomerization process. However, it is still
conformationally mobile. The figure thus shows the conformational isomerism for two different configurational
isomers, the enantiomeric pair.

A common mistake that can be made when one is trying to draw the different conformers that arise from ring flip
in a cyclohexane compound (see Box 3.3) is to remember vaguely that axial groups become equatorial, and vice
versa, and to apply this change without flipping the ring. Of course, as can be seen from looking at the compounds
below, transposing the equatorial bromine to axial and the axial methyl to equatorial changes the configuration
at both centres, so we have produced the enantiomer. This is a configurational isomer and not a conformer.

ax ax
H CH3 Br H

Br Br H H CH3 Br
eq eq

changing axial to equatorial and vice versa without
ring flip creates the enantiomer, not a conformer

90 STEREOCHEMISTRY

3.4.5 Meso compounds the substituent groups being the same, because these two
structures are actually superimposable and, therefore,
Now for a rather unexpected twist. We have seen only represent a single compound. This is not so easily
that if there are n chiral centres there should be 2n seen with the staggered conformers drawn, so it is best
to rotate these about the 2,3-bond to give an eclipsed
configurational isomers, and we have considered each conformer. They can both be rotated to give the same
of these for n = 2 (e.g. ephedrine, pseudoephedrine). It structure, so they represent only a single compound. This
transpires that if the groups around chiral centres are the is called meso-tartaric acid (Greek: mesos = middle).
same, then the number of stereoisomers is less than 2n. Furthermore, since we have superimposable mirror
Thus, when n = 2, there are only three stereoisomers, images, there can be no optical activity.
not four. As one of the simplest examples, let us consider
in detail tartaric acid, a component of grape juice and We can see why a compound with chiral centres
many other fruits. This fits the requirement, since each should end up optically inactive by looking again at
of the two chiral centres has the same substituents. the eclipsed conformer. The molecule itself has a plane
of symmetry, and because of this symmetry the optical
mirror activity conferred by one chiral centre is equal and
opposite to that conferred by the other and, therefore,
HO CO2H HO2C OH is cancelled out. It has the characteristics of a racemic
H H mixture, but as an intramolecular phenomenon. A meso
compound is defined as one that has chiral centres but
HO2C 23 H H 32 is itself achiral. Note that numbering is a problem in
OH HO CO2H tartaric acid because of the symmetry, and that positions
2 and 3 depend on which carboxyl is numbered as C-1. It
2S,3S 2R,3R can be seen that (2R,3S) could easily have been (3R,2S)
(+)-tartaric acid if we had numbered from the other end, a warning sign
(–)-tartaric acid that there is something unusual about this isomer.

mirror The same stereochemical principles apply to both
acyclic and cyclic compounds. With simple cyclic com-
H CO2H HO2C H pounds that have little or no conformational mobility, it
HO OH can even be easier to follow what is going on. Let us first
look at cyclopropane-1,2-dicarboxylic acid. These com-
HO2C 23 H H 32 pounds were considered in Section 3.4.3 as examples
OH HO CO2H of geometric isomers, and cis and trans isomers were
recognized.
2R,3S 2S,3R
meso-tartaric acid meso-tartaric acid

these two structures are
superimposable; this is more
easily seen by considering the
eclipsed conformer

plane of cyclopropane-1,2-dicarboxylic acid
symmetry
mirror

HO2C CO2H HO2C R H H S CO2H
R S
HH 1 1
HO OH 2 2
H H
meso-tartaric acid CO2H HO2C
eclipsed conformer
(+)- and (–)-trans enantiomers
because of the symmetry, optical activity
conferred by one chiral centre is equal and plane of
opposite to that conferred by the other; this symmetry
meso compound is optically inactive
HO2C R S CO2H

We can easily draw the four predicted isomers, as we HH
did for the ephedrine–pseudoephedrine group, and two
of these represent the enantiomeric pair of (−)-tartaric cis isomer is an optically
acid and (+)-tartaric acid. Now let us consider the other inactive meso compound
pair of isomers, and we shall see the consequences of
n = 2, but only three isomers

CONFIGURATIONAL ISOMERS 91

This is essentially the same as the tartaric acid cis-1,2-dimethylcyclohexane 1
example, without the conformational complication. 2
Thus, there are two chiral centres, and the groups around
each centre are the same. Again, we get only three CH3 mirror CH3
stereoisomers rather than four, since the cis compound
is an optically inactive meso compound. There is a plane CH3 H3C
of symmetry in this molecule, and it is easy to see that A C
one chiral centre is mirrored by the other, so that we
lose optical activity. CH3 CH3
D
Conformational mobility, such as we get in cyclo- CH3 H3C
hexane rings, makes the analysis more difficult, and B
manipulating molecular models provides the clear-
est vision of the relationships. Let us look at 1,2- this is the difficult one!
dimethylcyclohexane as an example. Again, we have
met the cis and trans isomers when we looked at con- the cis isomer is an optically inactive meso
formational aspects (see Section 3.3.2). Here, we need compound
to consider both configuration and conformation.
the picture shows mirror images of the
trans-1,2-dimethylcyclohexane 1 equal-energy interconvertible conformers
2
1 mirror 1 however, consider a 120° rotation of A about the
central axis which produces D; 120° rotation of C
CHC3H3 H3C produces B; therefore, they are all the same
H3C compound, but different conformers
2
2 superimposed. However, conformer A may be ring-
flipped to an equal-energy conformer B, and this will
CH3 H3C have a corresponding mirror image version D. Now
consider a 120◦ rotation of version A about the central
CH3 CH3 axis; this will give D. A similar 120◦ rotation of version
C about the central axis will give B. It follows, therefore,
(+)- and (–)- trans enantiomers; two chair that if simple rotation of one structure about its axis
conformations are shown for each, the favoured gives the mirror image of a conformational isomer,
diequatorial and the unfavoured diaxial − note that then we cannot have enantiomeric forms but must
the mirror image relationship is readily apparent in have the same compound. These are thus two different
both conformers conformers of an optically inactive meso compound. It
may require manipulation of models to really convince
Care: this shows two interconvertible conformers for you about this!
each of the two non-interconvertible enantiomers
Now, although the cyclohexane ring is not pla-
In the trans compound, two mirror image enan- nar, the overall consequences for trans- and cis-
tiomeric forms can be visualized. These will be the dimethylcyclohexane can be predicted by looking at the
(+)- and (−)-trans isomers. Note particularly that con- two-dimensional representations.
formational changes may also be considered, but these
do not change configuration, so we are only see- the meso nature of cis-1,2-dimethylcyclohexane
ing different conformers of the same compound. The can be deduced from the plane of symmetry in the
above scheme thus shows two interconvertible con- 2D representation:
formers (upper and lower structures) for each of the
two non-interconvertible enantiomers (left and right no plane of plane of
structures). symmetry symmetry

The cis compound provides the real challenge, trans cis
however. If we draw version A, together with its
mirror image C, they do not look capable of being

92 STEREOCHEMISTRY

It is clear that this representation of cis-dimethyl- • a molecule may be chiral without having a chiral
cyclohexane shows a plane of symmetry, and we centre.
can deduce it to be a meso compound. No such
plane of symmetry is present in the representation This is the subject of the next section.
of trans-dimethylcyclohexane. Why does this approach
work? Simply because the transformation of planar 3.4.6 Chirality without chiral centres
cyclohexane (with eclipsed bonds) into a non-planar
form (with staggered bonds) is a conformational change We shall restrict discussions here to three types of
achieved by rotation about single bonds. The fact that compound. In the first we get what is termed torsional
cyclohexane is non-planar means we may have to invoke asymmetry, where chirality arises because of restricted
the conformational mobility to get the three-dimensional rotation about single bonds. The commonest examples
picture. involve two aromatic rings bonded through a single
bond (biphenyls). If large groups are present in the
Our consideration of meso compounds leads us to ortho positions, these prevent rotation about the inter-
generalize: ring single bond, and the most favourable arrangement
to minimize interactions is when the aromatic rings
• a molecule with one chiral centre is chiral; are held at right angles to each other. As a result,
two enantiomeric forms of the molecule can exist.
• a molecule with more than one chiral centre may be Because of the size of the ortho groups, it is not
chiral or achiral. possible to interconvert these stereoisomers merely by
rotation. Even when we only have two different types of
Now let us extend this generalization with a further substituent, as shown, we get two enantiomeric forms.
statement:

CO2H mirror
Cl
HO2C rotate CO2H
Cl HO2C
structure 90º ≡
Cl
HO2C CO2H Cl
Cl Cl

large ortho groups prevent rotation
two enantiomeric forms exist

chirality via restricted rotation − torsional asymmetry

The second type of compound is called an allene; involved in two double bonds, it follows that the π
these compounds contain two double bonds involving bonds created must be at right angles to each other.
the same carbon. These compounds exist, but are often The consequence of this is that the substituents on the
difficult to prepare and are very reactive. It is the other carbons of the allene are also held at right angles
concept of chirality which is more important here than to each other. Again, two enantiomeric forms of the
the chemistry of the compounds. If a carbon atom is molecule can exist.

CONFIGURATIONAL ISOMERS 93

chirality in allenes overlap of p orbitals to generate
CCC π bonds means groups are held at
right angles to each other

mirror

H H rotate H Ph
H H structure CCC
≡
CCC CCC 90º H
Ph Ph Ph

Ph Ph

two enantiomeric forms without a chiral centre. Spiro compounds contain two
ring systems that have one carbon in common, and
The third example of chirality without a chiral centre it is easy to see this carbon could be chiral if four
is provided by spiro compounds, which we shall meet different groupings are present. A nice natural example,
later when we consider the stereochemistry of polycyclic the antibiotic griseofulvin, is shown here.
systems (see Section 3.5.1), but at this stage it is worth
noting that they provide a third example of chirality

spiro compounds OMe O OMe

spiro MeO * * chiral centre
rings share one atom O O

Cl this has a chiral centre
griseofulvin

mirror

NH HN rotate NH H
N N N
H H structure ≡ this has no chiral centre

90º

two enantiomeric forms

However, it is also possible to visualize spiro substituent are attached to the spiro centre. Both rings
compounds with groupings that are not all different, in this compound will have the chair conformation,
where enantiomeric forms exist because mirror image but it is not easy to draw these because one ring will
compounds are not superimposable. The diamine always be viewed face on. The solution is to ensure
shown is chiral, in that the mirror image forms are the spiro centre is not on the left or right tip of either
not superimposable, even though only two types of ring.

it is difficult to show the chair the solution is to ensure
conformation for both rings the spiro centre is not on
the tip of either ring
rotate
HN structure H NH
N
HN
NH
NH

94 STEREOCHEMISTRY

With biphenyls, allenes, and spiro compounds, groups A A
are held at right angles by a rigid system, and this feature DD
allows the existence of non-superimposable mirror image EE
stereoisomers, i.e. enantiomers. It is useful to think of this B B
arrangement as analogous to a simple chiral centre, where
the tetrahedral array also holds pairs of groups at right with biphenyls, allenes, and spiro compounds,
angles. In contrast to tetrahedral carbon, it is not even groups are held at right angles by a rigid system;
necessary for all the groups to be different to achieve the arrangement produces non-superimposable
chirality, as can be seen in the examples above. mirror images and is thus analogous to a
chiral centre

Box 3.12

Torsional asymmetry: gossypol

The concept of torsional asymmetry is not just an interesting abstract idea. Some years ago, fertility in some Chinese
rural communities was found to be below normal levels, and this was traced back to the presence of gossypol in
dietary cottonseed oil. Gossypol acts as a male contraceptive, altering sperm maturation, spermatozoid motility,
and inactivation of sperm enzymes necessary for fertilization. Extensive trials in China have shown the antifertility
effect is reversible after stopping the treatment, and it has potential, therefore, as a contraceptive for men.

OHC OH HO CHO
HO OH HO OH

HO OH HO OH
HO CHO OHC OH

(+)-gossypol (–)-gossypol

Gossypol is chiral due to restricted rotation, and only the (−)-isomer is pharmacologically active as an
infertility agent. The (+)-isomer has been found to be responsible for some toxic symptoms. Most species of
cotton (Gossypium) produce both enantiomers of gossypol in unequal amounts, with the (+)-enantiomer normally
predominating over the (−)-isomer. It has proved possible to separate racemic (±)-gossypol from this type of
mixture – the racemate complexes with acetic acid, whereas the separate enantiomers do not. The racemic form
can then be resolved (see Section 3.4.8) to give the useful biologically active (−)-isomer.

3.4.7 Prochirality mirror mirror
AA AA
Enantiotopic groups
E B B EA BB A
We have defined chirality in terms of ‘handedness’, such D DD D
that mirror image stereoisomers are not superimposable.
In the case of tetrahedral carbon, chirality is a conse- chiral centre Molecules that are
quence of having four different groups attached to it. superimposable on their
If two or more groups were the same, then the com- mirror images are
pound would be termed achiral (see Section 3.4.1). Now achiral
we introduce another term, prochiral. Achiral molecules
that can become chiral by one simple change are called achiral molecules, that can become chiral
prochiral. The simplest example we could include under by one simple change are called prochiral;
this definition would be an achiral molecule in which the A groups are termed enantiotopic
two groups are the same. The two like groups are termed
enantiotopic, in that separate replacement of each would A EA
generate enantiomers.
A B A or E B
D D B D

enantiomers

CONFIGURATIONAL ISOMERS 95

This seems an unnecessary complication. Why do we use pro-R and pro-S descriptors
want to call an achiral centre prochiral? What benefits to distinguish enantiotopic
are there? Well, remember that the Cahn–Ingold–Prelog hydrogens/groups
system allowed us to describe a particular chiral
arrangement of groups at a chiral centre; prochirality pro-S pro-R HS HR
now allows us to distinguish between the two like groups HH
at an achiral centre. When might we want to do that? The
following example from biochemistry shows the type of H3C OH H3C OH
occasion when we might need to identify one or other
of the like groups. pro-R HD
HH
The enzyme alcohol dehydrogenase oxidizes ethanol
to acetaldehyde, passing the hydrogen to the coen- H3C OH H3C R OH
zyme nicotinamide adenine dinucleotide NAD+ (see
Section 15.1.1). This is the enzyme that restores normal increasing the priority of
service after excessive consumption of alcoholic drinks. the pro-R hydrogen
By specifically labelling each hydrogen in turn, then creates R configuration
observing whether the substrate loses or retains label
in the enzymic reaction, it has been determined which pro-S DH
hydrogen is lost from the methylene group of ethanol. HH

H3C OH H3C S OH

HH increasing the priority of
the pro-S hydrogen
H3C OH creates S configuration
ethanol is prochiral

alcohol O HD alcohol O
H D dehydrogenase H3C H dehydrogenase H3C H

H3C R OH NAD+ H3C R OH NAD+

alcohol the enzyme is stereospecific;
D H dehydrogenase it removes the pro-R hydrogen

O

H3C S OH NAD+ H3C D This example is from biochemistry. It is a feature
of biochemical reactions that enzymes almost always
How then, in unambiguous fashion, can we describe catalyse reactions in a completely stereospecific manner.
which hydrogen is lost? We define the two hydrogens They are able to distinguish between enantiotopic
as pro-R and pro-S, by considering the effect of hydrogens because of the three-dimensional nature of
increasing their effective priorities according to the the binding site (see Section 13.3.2). There are also
Cahn–Ingold–Prelog system; this is simply achieved occasions where chemical reactions are stereospecific;
if we consider having deuterium instead of protium refer to the stereochemistry of E2 eliminations for
(normal hydrogen). Then, if replacing a particular typical examples (see Section 6.4.1).
hydrogen with deuterium produces a chiral centre with
the R configuration, that hydrogen is termed the pro-R Box 3.13
hydrogen. Similarly, increasing the priority of the other
hydrogen should generate the S configuration, so that Citric acid has three prochiral centres
that hydrogen is termed the pro-S hydrogen. We can also
label hydrogens in a structure as HR and HS according The Krebs cycle is a process involved in
to this procedure. the metabolic degradation of carbohydrate (see
Section 15.3). It is also called the citric acid cycle,
We can thus deduce that alcohol dehydrogenase because citric acid was one of the first intermediates
stereospecifically removes the pro-R hydrogen from the identified. Once formed, citric acid is modified
prochiral methylene. by the enzyme aconitase through the intermediate

96 STEREOCHEMISTRY

cis-aconitic acid to give the isomeric isocitric acid. can only label this hydrogen as pro-S in citric acid;
This is not really an isomerization, but the result in cis-aconitic acid and isocitric acid, it is no longer
of a dehydration followed by a rehydration. Both attached to a prochiral centre, and we must resort to
steps feature stereospecific anti processes, i.e. groups some other labelling system, namely the asterisk.
are removed or added from opposite sides of the
molecule (see Sections 6.4.1 and 8.1.2). This is a nice example of enzymic stereospeci-
ficity. It involves specific removal of one hydrogen
citric acid has three prochiral centres; atom from a substrate that appears to have four equiv-
it is also prochiral at the central carbon alent hydrogens. Because of the three-dimensional
characteristics of both the enzyme and the substrate,
HO CO2H the apparently equivalent side-chains on the central
carbon are going to be positioned quite differently
HO2C CO2H and the enzyme is able to distinguish between them.
pro-S pro-R Further, it also distinguishes between the two hydro-
gens of a methylene group. An interesting conse-
*HS HR quence of this stereospecificity is that, because only
one of the citric acid side-chains is modified in the
−H2O aconitase aconitase reaction, it takes further turns of the cycle
anti-elimination before material entering the cycle (acetyl-CoA) is
actually degraded (see Section 15.3).
HO2C CO2H
CO2H A reaction that gives a mixture of isomeric
products with one isomer predominating would be
termed stereoselective.

H* note:
cis-aconitic acid H* not HS

Enantiotopic faces

+ H2O aconitase We have thus seen that there could be a need to distin-
anti-addition guish between two similar groups attached to tetrahedral
carbon, and have exploited the Cahn–Ingold–Prelog
H CO2H priorities to label the separate groups. We also need to
consider another way in which a chiral centre might be
HO2C CO2H generated, and that is by addition of a group to a pla-
nar system. For example, if we reduce a simple ketone
*H OH note: that has two different R groups with lithium aluminium
isocitric acid H* not HS hydride we shall produce a racemic alcohol product (see
Section 7.5). This is because hydride can be delivered
aconitase removes the pro-R hydrogen to either face of the planar carbonyl group with equal
from the pro-R substituent probability.

First, let us look closely at the structure of citric addition from either face of
acid. It has three prochiral centres. Two of these are planar carbonyl group
the methylenes, but note that the central carbon is
also prochiral. It has two groups the same, namely R´ LiAlH4 R´ R´
the –CH2CO2H groups. The loss of water from citric
acid is an anti elimination, so that the hydroxyl is O R OH + H OH
lost together with one of the methylene hydrogens. H R
The hydrogen lost has been found to be the pro-R R
hydrogen from the pro-R–CH2CO2H group.
In marked contrast, nature’s reducing agent, reduced
This is followed by an anti addition reaction in nicotinamide adenine dinucleotide (NADH), delivers
which water is added to the new double bond, but in hydride in a stereospecific manner because it is a
the reverse sense. The hydrogen retained throughout cofactor in an enzyme-catalysed reaction. For example,
the process is shown with an asterisk. Note that we reduction of pyruvic acid to lactic acid in vertebrate
muscle occurs via attack of hydride to produce just one
enantiomer, namely (S)-lactic acid.

CONFIGURATIONAL ISOMERS 97

O NADH HO H stereospecific reduction;
hydride delivered to front
H3C CO2H lactate H3C CO2H face (Re)
(S)-(+)-lactic acid
pyruvic acid dehydrogenase

We can see from the diagram that hydride must be front or back. Once again, the Cahn–Ingold–Prelog
delivered from the front face as shown, but it makes system can help us out. We assign priorities to the three
sense to have a more precise descriptor for faces than groups attached to the planar carbon.

priority 1 clockwise: Re priority 1 anticlockwise: Si
O O

priority 3 H3C CO2H priority 2 priority 2 HO2C CH3 priority 3
pyruvic acid pyruvic acid

We then consider the descending sequence and decide Note that there is no correlation between Re or Si
whether this is clockwise or anticlockwise; the face and the chirality R or S of the tetrahedral product
that provides a clockwise sequence is then labelled Re formed.
and the face that provides an anticlockwise sequence
is labelled Si. These are simply variants on R and It can now be seen that, in the enzymic reduction of
S, in fact the first two letters of rectus and sinister. pyruvic acid to lactic acid, hydride is delivered to the
Re face of the pyruvic acid.

top face stereospecific reduction;
hydride delivered to Re face
H

H3C Re face H3C CO H OH
HO2C CO HO2C H
Si face H3C
HO2C

bottom face pyruvic acid (S)-(+)-lactic acid

A molecule such as pyruvic acid is said to have two a little more complex, in that a C=C bond generates
enantiotopic faces. Attack of a reagent onto the Re face four faces to be considered, two at each carbon. It is
yields one enantiomer, whereas attack onto the Si face necessary to systematically deduce the descriptor for
will produce the other enantiomer. each, as shown below.

The Re and Si descriptors are similarly applied to
the carbon atoms making up C=C bonds. This gets

priority 2 priority 3

H3C •3 H Re

priority 1

H3C 3 H H3C CO2H Si Re Si Si

H3C 2 CO2H H3C H H3C CC CH3 H3C CC H
priority 2 HO2C H HO2C CH3
23 23
• Si
H3C 2 CO2H Re Si Re Re
priority 3 priority 1
each sp2 carbon has two faces

98 STEREOCHEMISTRY

Box 3.14

NADH delivers hydride from a prochiral centre; NAD+ has enantiotopic faces

NADH (reduced nicotinamide adenine dinucleotide) is utilized in biological reductions to deliver hydride to an
aldehyde or ketone carbonyl group (see Box 7.6). A proton from water is used to complete the process, and the
product is thus an alcohol. The reaction is catalysed by an enzyme called a dehydrogenase. The reverse reaction
may also be catalysed by the enzyme, namely the oxidation of an alcohol to an aldehyde or ketone. It is this
reverse reaction that provides the dehydrogenase nomenclature.

During the reduction sequence, NADH transfers a hydride from a prochiral centre on the dihydropyridine ring,
and is itself oxidized to NAD+ (nicotinamide adenine dinucleotide) that contains a planar pyridinium ring. In the
oxidation sequence, NAD+ is reduced to NADH by acquiring hydride to an enantiotopic face of the planar ring.
The reactions are completely stereospecific.

biological reduction−oxidation via hydride transfer

H3C H3C H
CO H HCO

H reduction of ketone H

pro-S pro-R HH alcohol H Re face
HH CONH2 dehydrogenase CONH2
4 CONH2 H
RN CONH2
Si face
NN N

R R oxidation of alcohol R

NADH NAD+

nicotinamide adenine nicotinamide adenine

dinucleotide (reduced) dinucleotide

reducing agent; can oxidizing agent;
supply hydride can remove hydride

The stereospecificity depends upon the enzyme in question. Let us consider the enzyme alcohol dehydrogenase,
which is involved in the ethanol to acetaldehyde interconversion. It has been deduced that the hydrogen transferred
from ethanol is directed to the Re face of NAD+, giving NADH with the 4R configuration. In the reverse reaction,
it is the 4-pro-R hydrogen of NADH that is transferred to acetaldehyde.

Note also that transfer of hydride to the carbonyl compound is also stereospecific, as is removal of hydrogen
from the prochiral centre of ethanol in the reverse reaction (see Section 3.4.7).

We should note that prochiral molecules have the production of enantiomers. However, if there is also a
potential to become chiral if we make certain changes, chiral centre in the molecule, then the same changes
and we have used the term enantiotopic to identify the would lead to the formation of diastereoisomers, not
groups at sp3-hybridized carbon or the faces of sp2- enantiomers. Such groups or faces are now correctly
hybridized carbon where alternative changes lead to the termed diastereotopic.

enantiotopic diastereotopic
hydrogens hydrogens

HH HH O chiral centre
H3C OH H3C OH
H3C H O
H OH H3C H
molecule has
chiral centre enantiotopic faces H OH
molecule has
diastereotopic faces

CONFIGURATIONAL ISOMERS 99

3.4.8 Separation of enantiomers: resolution The bases generally employed in such resolutions
have been natural alkaloids, such as strychnine, brucine,
We saw in Section 3.4.1 that enantiomers have the and ephedrine. These alkaloids are more complex than
same physical and chemical properties, except for opti- the general case shown in the figure, in that they
cal activity, and thus they behave in exactly the same contain several chiral centres (ephedrine is shown in
manner. We also saw, however, that this generalization Section 3.4.4). Tartaric acid (see Section 3.4.5) has been
did not extend into biological properties, and that there used as an optically active acid to separate racemic
were compelling reasons for administering drugs as a bases. Of course, not all materials contain acidic or
single enantiomer rather than a racemate (see Box 3.7). basic groups that would lend themselves to this type of
At some stage, therefore, it might be necessary to have resolution. There are ways of introducing such groups,
the means of separating individual enantiomers from a however, and a rather neat one is shown here.
racemic mixture. This is termed resolution. The tra-
ditional method has been to convert enantiomers into A
diastereoisomers, because diastereoisomers have differ-
ent physical and chemical properties and can, therefore, O HO
be separated by various methods (see Section 3.4.4). O B
Provided one can convert the separated diastereoisomers O
back to the original compound, this offers a means of phthalic anhydride C
separating or resolving enantiomers.
+ ester
The simplest method has been to exploit salt for- A
mation by reaction of a racemic acid (or base) with a formation
chiral base (or acid). For example, treating a racemic HO
acid with a chiral base will give a mixture of two salts B
that are diastereoisomeric. Although there is no covalent C
bonding between the acid and base, the ionic bond-
ing is sufficient that the diastereoisomeric salts can be racemic alcohol
separated by some means, typically fractional crystal-
lization. Although fractional crystallization may have to CO2H A
be repeated several times, and, therefore, is tedious, it O B
has generally been an effective means of separating the
diastereoisomeric salts. Finally, the salts can separately O+ C
be converted back to the acid, completing the resolution.
salt CO2H A
formation O B

diastereoisomeric O C
salts
HO2C AX NH3 O2C A separate racemic acid
+ BY + B
X regenerate free
Y NH2 HO2C CZ NH3 O2C C acid
salt
Z A hydrolyse ester
A formation X B
optically BY
active base C
CZ

racemic acid pair of diastereoisomeric
salts
separated
A regenerate X A enantiomeric
free acid
HO2C Y NH3 O2C B alcohols
B ZC
C H+ A racemic alcohol may be converted into a racemic
acid by reaction with one molar equivalent of phthalic
HO2C A H+ X NH3 O2C A separate anhydride; the product is a half ester of a dicarboxylic
B Y physically acid (see Section 7.9.1). This can now be subjected to
regenerate the resolution process for acids and, in due course, the
C free acid Z B alcohols can be regenerated by hydrolysis of the ester.
C
A significant improvement on the fractional crystal-
separated lization process came with the introduction of chiral
enantiomers

100 STEREOCHEMISTRY

phases for column chromatography. This allows sim- O lipase A
ple chromatographic separation of enantiomers. In prac- Me A HO
tice it is effectively the same principle, that of forming
diastereoisomeric complexes with the chiral material O B
comprising the column. One enantiomer binds more B C
tightly than the other and, therefore, passes through the +
column at a different rate. The two enantiomers thus C
emerge from the column as separate fractions. + O
Me A
It has also proved possible to exploit the enantiospe- O
cific properties of enzymes to achieve resolution of a Me A O
racemic mixture during a chemical synthesis. Enzymes B
(see Section 13.4) are proteins that catalyse biochemi- O
cal reactions with outstanding efficiency and selectivity. B C
This is a consequence of the size and shape of the
enzyme’s binding site, a feature that is determined by C only one enantiomer
the sequence of amino acid residues in the protein (see is hydrolysed
Section 13.3.2). The selectivity of enzymes means that racemic ester
they carry out reactions on one functional group in the
presence of others that might be affected by a chemical separate
reagent. It also means that they can be stereoselective,
either performing reactions in a stereospecific manner or A
only reacting with substrates with a particular chirality. HO
As a simple example, racemic ester structures may be
resolved by the use of ester hydrolysing enzymes called B
lipases. C

With the appropriate choice of enzyme, it has been A base O
found that only one enantiomer of the racemic mixture HO Me A
is hydrolysed, whilst the other remains unreacted. It is
then a simple matter to separate the unreacted ester B O
from the alcohol. The unreacted ester may then be C B
hydrolysed chemically, thus achieving resolution of the
enantiomeric alcohols. C

3.4.9 Fischer projections Section 12.2). To start, though, let us consider just one
chiral centre, and choose the amino acid we met earlier
Fischer projections provide a further approach to the (see Section 3.4.2), (−)-(S)-serine.
two-dimensional representations of three-dimensional
formulae. They become particularly useful for molecules The Fischer projection is drawn with groups on
that contain several chiral centres, and are most horizontal and vertical lines, but without showing the
frequently encountered in discussions of sugars (see chiral carbon atom. Should you put in this carbon
atom, it can no longer be considered that you are
representing stereochemistry. The Fischer projection
then implies that horizontal bonds are wedged, whilst
vertical bonds are dotted, and it thus speeds up the
drawing of stereochemical features. For (−)-(S)-serine,
the wedge–dot version is what one would see if
one looked down on the right-hand stereostructure

Fischer projection carbon with highest Fischer projection is equivalent to
oxidation state at top viewing molecule from the top

CO2H ≡ CO2H ≡ H2N H
H2N H HOH2C CO2H
H2N H
(–)-(S)-serine
carbon not shown − CH2OH CH2OH
intersection of lines

longest carbon horizontal lines above plane
chain vertical vertical lines below plane

CONFIGURATIONAL ISOMERS 101

as indicated. Accordingly, we can now transform tions, we may need to disregard these restrictions in the
stereostructures into Fischer projections, and vice versa. interests of following the changes.

The only significant restrictions are Manipulations we can do to a Fischer projection
may at first glance appear confusing, but by reference
• we should draw the longest carbon chain vertical; to a model of a tetrahedral array, or even a sketch
of the representation, they should soon become quite
• we should place the carbon of highest oxidation state understandable, perhaps even obvious. The molecular
at the top. manipulations shown are given to convince you of the
reality of the following statements.
However, when we come to manipulate Fischer projec-

• Rotation of the formula by 180◦ gives the same • Rotation of any three groups clockwise or anticlock-
molecule. wise gives the same molecule.

A 180º C AB
D B ≡B D D B≡D C

CA CA

A B≡ 180º B D≡ C A B≡ DB D B
D AC BD D CA ≡ C
D B≡
C CA A C ≡D C

AB A

• Exchange of any two groups gives the enantiomer.

original isomer

A mirror
DC
A A DB DC CD DB A
DB B DB ≡ ≡
enantiomer BA ≡D B
C C CA enantiomer AB
AC C

exchange of two mirror image of
groups gives enantiomer is
enantiomer original isomer

• Rotation of the formula by 90◦ gives the enantiomer.

original isomer

mirror

A 90º D A DB BD CA D
DB ≡ CA ≡≡ CA
DB CA
C CA BD B
CB

enantiomer enantiomer one exchange gives
enantiomer, second
exchange restores
original isomer

It is also surprisingly easy to assign R or S configu- • if the group of lowest priority is on the vertical line,
rations to chiral carbons in the Fischer projections; but, a clockwise sequence gives the R configuration;
because horizontal lines imply wedged bonds (towards
you) and vertical lines imply dotted bonds (away from • if the group of lowest priority is on the horizontal
you), there are important guidelines to remember: line, a clockwise sequence gives the S configuration.

102 STEREOCHEMISTRY

These do not represent a different set of rules from horizontal line. We have noted (see Box 3.8) that, if the
the clockwise = R, anticlockwise = S conventions we lowest priority group is wedged, it is easier to look at
already use (see Section 3.4.2). It is merely a conse- the sequence from the front, then reverse it to give us
quence of the lowest priority group being down (dot- the sequence as viewed from the rear, i.e. towards the
ted bond) on the vertical line, but up (wedged) on the group of lowest priority.

rotation by 180º
gives same molecule

R if group of lowest priority is on the vertical line,
1 180º 4 a clockwise sequence gives the R configuration
3 2 ≡2 3
if group of lowest priority is on the horizontal line,
interchange of 4 R1 a clockwise sequence gives the S configuration
two groups gives S
enantiomer
1

42 relate this to horizontal bonds implying wedged (up)
and vertical bonds implying dotted (down)
3

numbers refer to H H S
assigned priorities R

hydrogen down, hydrogen up, must view from rear;
clockwise = R alternatively, front view clockwise
needs reversing = S

Let us apply these principles to tartaric acid. This is an optically inactive meso compound involved (see
compound has two chiral centres; but, as we saw Section 3.4.5).
previously, only three stereoisomers exist, since there

CO2H 2 2 CO2H CO2H
H 2 OH
H2 OH R 1 1S HO 2 H H 3 OH
HO 3 3 3
3 H 3 OH 3 CO2H plane of
1R H S1 symmetry

2 CO2H CO2H 2

H on horizontal H on horizontal (2R,3S)-meso-tartaric acid
anticlockwise = R clockwise = S

(2R,3R)-(+)-tartaric acid (2S,3S)-(–)-tartaric acid

We can draw these three stereoisomers as Fischer line is the longest carbon chain. Thus, we have to
projections, reversing the configurations at both centres reverse our normal configurational thinking: a clockwise
to get the enantiomeric stereoisomers, whilst the Fischer sequence of priorities gives S and an anticlockwise
projection for the third isomer, the meso compound, is sequence gives R. The configuration of the meso isomer
characterized immediately by a plane of symmetry. For can be deduced by abstracting the appropriate portions
(+)-tartaric acid, the configuration is (2R,3R), and for from the other two structures and assigning equivalent
(−)-tartaric acid it is (2S,3S). For both chiral centres, configurations.
the group of lowest priority is hydrogen, which is on
a horizontal line. In fact, this is the case in almost all It should be appreciated that a Fischer projection
Fischer projections, since, by convention, the vertical involving more than one chiral centre actually depicts
an eclipsed conformer, which is naturally a high-energy

CONFIGURATIONAL ISOMERS 103

state, and is normally an unlikely arrangement of atoms stereochemical drawings, with further manipulations
(see Section 3.3.1). We need to bear this in mind necessary to give lower energy staggered conformers. This
when we transpose Fischer projections into wedge–dot is illustrated here with the five-carbon sugar (−)-ribose.

Fischer projection is equivalent
to viewing the eclipsed conformer
from the top

CHO CHO H OH H OH
H OH HH
H OH H OH HO OH ≡ HOH2C CHO
H OH OH HOCH2 CHO
≡≡ H
CH2OH eclipsed conformer
H OH H OH H OH

CH2OH staggered conformer

(–)-ribose implied
Fischer projection stereochemical

relationship

However, as we shall see shortly, Fischer-projection- CHO CHO
derived eclipsed conformers are particularly useful in H OH HO H
deducing the stereochemistry in cyclic forms of sugars
(see Box 3.16). CH2OH CH2OH

3.4.10 D and L configurations (R)-(+)-glyceraldehyde (S)-(–)-glyceraldehyde
= =
The concept of D and L as configurational descriptors is
well established, particularly in amino acids and sugars; D-(+)-glyceraldehyde L-(–)-glyceraldehyde
frankly, however, we could live without them and save
ourselves a lot of confusion. Since they are so widely 1 CHO 1 CHO
used, we need to find out what they mean, but in most H 2 OH HO 2 H
cases the information conveyed is less valuable than HO 3 H
sticking with R and S. H 4 OH H 3 OH
H 5 OH HO 4 H
D and L sugars HO 5 H
6 CH2OH
The simplest of the sugars is glyceraldehyde, which has 6 CH2OH
one chiral centre. Long before R and S were adopted D-(+)-glucose
as descriptors, the two enantiomers of glyceraldehyde L-(–)-glucose
were designated as D and L. D-(+)-Glyceraldehyde is
equivalent to (R)-(+)-glyceraldehyde, the latter config- Since the configuration at position 5 in (+)-glucose
uration being fully systematic. Configurations in other can be directly related to that in D-(+)-glyceraldehyde,
compounds were then related to the configurations of D- (+)-glucose is said to have the D configuration, and is
and L-glyceraldehyde by direct comparison of Fischer thus termed D-(+)-glucose. By similar reasoning, the
projections. For example, (+)-glucose (= dextrose) is enantiomer of glucose has the L configuration, and is
represented by a Fischer projection that defines the con- termed L-(−)-glucose. Now the limitations of this system
figuration at all four chiral centres. become obvious when one realizes that D and L refer
to the configuration at just one centre, by convention
the highest numbered chiral centre, and the remaining
configurations are not specified, except by the name of
the sugar (see Box 3.15).

104 STEREOCHEMISTRY

Box 3.15

Fischer projections of glucose and stereoisomers

The sugar glucose has four chiral centres; therefore, 24 = 16 different stereoisomers of this structure may be
considered. These are shown below as Fischer projections.

CHO CHO 1CHO CHO CHO CHO CHO CHO
H OH HO H H 2 OH HO H
H OH HO H HO 3 H HO H H OH H OH HO H
H OH H 4 OH H OH
H OH H OH H 5 OH HO H H OH HO H HO H HO H
H OH
CH2OH 6 CH2OH H OH HO H H OH HO H HO H
D-(+)-allose CH2OH CH2OH
D-(+)-altrose D-(+)-glucose H OH H OH H OH H OH
D-(+)-idose
CH2OH CH2OH CH2OH CH2OH

D-(+)-mannose D-(–)-gulose D-(+)-galactose D-(+)-talose

(C-3 epimer of (C-2 epimer of (C-4 epimer of
D-glucose) D-glucose) D-glucose)

CHO CHO 1 CHO CHO CHO CHO CHO CHO
HO H H OH HO 2 H H OH
HO H H OH H OH HO H HO H HO H H OH
HO H HO H H 3 OH H OH
HO H HO H HO 4 H H OH HO H HO H H OH H OH
HO 5 H
CH2OH CH2OH HO H H OH CH2OH H OH H OH
L-(–)-allose L-(–)-altrose 6 CH2OH L-(–)-idose
HO H HO H HO H HO H
L-(–)-glucose
CH2OH CH2OH CH2OH CH2OH

L-(–)-mannose L-(+)-gulose L-(–)-galactose L-(–)-talose

(C-5 epimer of
D-glucose)

The 16 stereoisomers are divided into D and L groups, which reflect only the configuration at the highest numbered
chiral centre, namely C-5. The chirality at other centres is defined solely by the name given to the sugar, so we
have eight different names for particular configurational combinations. Note that although D and L strictly refer
to the configuration at only one centre, L-glucose is the enantiomer of D-glucose and, therefore, must have the
opposite configuration at all chiral centres. A change in configuration at only one centre produces a diastereoisomer
that has different chemical properties, and is accordingly given a different name.

Whilst this system of nomenclature has some obvious shortcomings, it is analogous to the ephedrine and
pseudoephedrine example where we were considering just two chiral centres (see Section 3.4.4). A more
systematic approach (though not one that is used) might give all the above sugars the same name, e.g. hexose, but
specify the chirality at each centre, e.g. D-(+)-glucose would be (+)-(2R,3S,4R,5R)-hexose and L-(−)-galactose
would become (−)-(2S,3R,4R,5S)-hexose. Instead, we have the eight different names in two configurational
classes, D and L.

We can also use the term epimer to describe the relationship between isomers, where the difference is in the
configuration at just one centre (see Section 3.4.4). This is shown for the four epimers of D-(+)-glucose. An
interesting observation with the 16 stereoisomers is that optical activity of a particular isomer does not appear to
relate to the configuration at any particular chiral centre.

Box 3.16

Stereochemistry in hemiacetal forms of sugars from Fischer projections

In solution, aldehyde sugars normally exist as cyclic hemiacetals through reaction of one of the hydroxyls with
the aldehyde group, giving a strain-free six- or five-membered ring (see Section 3.3.2). The Fischer projection
for the sugar is surprisingly useful in predicting the configuration and conformation of the cyclic form.

CONFIGURATIONAL ISOMERS 105

rotate three groups; this brings the

5-hydroxyl onto the main chain which

will then form the ring system

'up' substituents shown in bold

1 CHO rotate CHO CHO
OH
H2 OH three H H H OH turn 6 H
HO 3 groups HO OH HOH2C H OH H O
H ≡ HO H 90º formation of
H4 H≡ H OH HO 5 4 3 2 1 hemiacetal
H ≡ H OH H OH H (see Section 7.2)
OH HOH2C

5 HOH2C H

H OH

6 CH2OH OH OH

D-(+)-glucose 6

H 6 CH2OH 'up' substituents
shown in bold
4 CH2OHHO H 5 O OH

HO H
HO 5 4 OH H 1
2 1 OH
3 OH 3 2H
cyclic hemiacetal OH
HH H H OH
form of D-(+)-glucose

this conformational drawing is the commonly used Haworth
much more informative than representation follows directly
the Haworth representation from the Fischer projection

The approach is straightforward. Since cyclic hemiacetal formation requires a hydroxyl group as the nucleophile
to attack the protonated carbonyl (see Section 7.2), we put this hydroxyl group on the vertical, thus getting all
the ring atoms onto the vertical. This requires rotation of three groups attached to the appropriate atom, C-5 in
the case of D-(+)-glucose. Such rotation does not affect the configuration at C-5. Then put in the stereochemistry
implied by the Fischer projection, using wedges and dots. This structure should then be turned on its side, and
the ring formation considered by joining up the C-5 hydroxyl and the carbonyl at the rear of the structure. Note
that, as drawn, this eclipsed conformer from the Fischer projection actually has these atoms quite close together,
so that ring formation is easily achieved and, most importantly, easily visualized (see Section 3.4.9).

The net result is a cyclic system looking like the Haworth representation that is commonly used, especially
in biochemistry books. The Haworth representation nicely reflects the up–down relationships of the various
substituent groups, but is uninformative about whether these are equatorial or axial. The last step, therefore, is to
transcribe this representation into a chair conformation, as shown, so that we see the conformational consequences.

rotate 'up' substituents shown in bold
three
1 CHO groups CHO CHO turn
H 2 OH OH
H 3 OH ≡ H OH H OH 90º 5 H
H 4 OH H H OH
≡H ≡ HOH2C H H O formation of
5CH2OH OH hemiacetal
HOH2C HOH2C H HO 4 3 2 1

OH H OH OH H

D-(−)-ribose 5

5 HOH2C H O H OH

HOH2C H O H OH 41

41

cyclic hemiacetal H3 2H H3 2H
HO OH OH OH

form of D-(−)-ribose

more informative Haworth representation

conformational drawing

106 STEREOCHEMISTRY

The alternative chair conformation, should we draw it instead, would be less favoured than that shown because
of the increased number of axial substituents. The conformation of D-glucose is the easily remembered one, in
that all the substituents are equatorial.

A similar procedure is shown for D-(−)-ribose, which, although it is capable of forming a six-membered cyclic
form, is found to exist predominantly as a five-membered ring (see Section 12.2.2).

D and L amino acids the integrity of the chiral centre. The fine detail of
the transformations need not concern us here. The net
There is a correlation between D- and L-glyceraldehyde result is that D- and L-amino acids have the general
and D- and L-amino acids, in that it is possible to convert configurations shown.
one system chemically into another without affecting

exchange NH2/R

exchange H/CO2H view from

CO2H CO2H H top
H NH2
H2N H ≡R CO2H ≡ R CO2H
R
D-amino acids R NH2 H2N H

L-amino acids L-amino acids

the common way of
presenting L-amino acids

Note that all the amino acids found in proteins are 3.5 Polycyclic systems
of the L configuration (excepting the achiral glycine);
D-amino acids are found in some polypeptide antibiotics Many molecules of biological or pharmaceutical impor-
(see Section 13.1). As we pointed out in Section 3.4.2, tance contain polycyclic ring systems, and we have
this brings up an apparent anomaly in nomenclature. already met some examples in other contexts, e.g. peni-
In all protein L-amino acids, except for cysteine, this cillins (see Box 3.8). There are three main ways in
represents an S configuration; cysteine, because of its which rings can be joined together, according to whether
high-priority sulfur atom has the R configuration. One they share one atom, two atoms, or more than two atoms.
can consider they all have the same configuration based These are termed spiro, fused, or bridged systems
on the L descriptor, but the priority rules lead to a respectively. Examples are shown where six-membered
different label. rings are joined in the various ways, but the concepts
apply equally to rings of other sizes.
One further point; as mentioned in Section 3.4.1, the
now obsolete descriptors d and l are abbreviations for
dextrorotatory (+) and laevorotatory (−) respectively.
They do not in any way relate to D and L.

except for L-cysteine, L-cysteine is R spiro fused
all the L-amino acids in proteins share 1 atom share 2 atoms
have the S configuration;

H on horizontal L-cysteine
clockwise is S
CO2H 3 bridged
CO2H 2 share >2 atoms
H2N H 1 S H2N H 1 R
3.5.1 Spiro systems
R3 CH2SH 2
Spiro systems have two rings sharing a single carbon
Priorities Priorities atom, and since this has essentially a tetrahedral array of

NH2 > CO2H > Alkyl > H NH2 > CH2SH > CO2H > H

CONFIGURATIONAL ISOMERS 107

bonds, the bonds starting the two rings must be arranged Solasodine and tomatidine are steroidal alka-
perpendicular to each other. If there is appropriate loids produced by potatoes (Solanum tuberosum)
substitution on the rings, then this can lead to the spiro and tomatoes (Lycopersicon esculente) respectively.
centre becoming chiral (see Section 3.4.6). These compounds, as glycosides (see Section 12.4),
are responsible for the toxic properties of the foliage
Box 3.17 and green fruits of these plants. They are not present
in potato tubers, unless green, or in ripe tomato fruits.
Natural spiro compounds Both compounds contain a spiro system, a nitrogen
analogue of a ketal (see Section 7.2). A spiroketal
Spiro compounds are exemplified by several natural is present in diosgenin from Dioscorea species, a
product structures. One of these is the antifungal raw material used for the semi-synthesis of steroidal
agent griseofulvin produced by cultures of the mould drugs. Note that solasodine and tomatidine demon-
Penicillium griseofulvum. Griseofulvin is the drug of strate the different configurations at the spiro centre;
choice for many fungal infections, but it is ineffective all natural spiroketals have the same stereochemistry
when applied topically, so is administered orally. at the spiro centre as in diosgenin.
Griseofulvin has two chiral centres, one of which
is the spiro centre, so there are potentially four 3.5.2 Fused ring systems
configurational isomers for the structure. Natural
griseofulvin has the configurations shown.

OMe O OMe Fused ring systems are particularly common. It is logical
to suppose that fusing on one or more additional ring
O systems is going to have stereochemical consequences,
in particular that the conformational changes seen
MeO O with single ring systems are likely to be significantly
modified. Initially, let us consider two cyclohexane
Cl rings fused together, giving a bicyclic system called
griseofulvin decalin.

H trans ring fusion H
N H
H

H O ≡
HH H

HO solasodine H H
trans-decalin

H NH cis ring fusion
H
O H
H
H ≡
HH

HO H H
H cis-decalin
tomatidine

O Two configurational isomers exist, trans- and cis-
decalin, according to the stereochemistry of ring fusion.
H The trans or cis relationship is most easily seen with the
hydrogens at the ring fusion carbons, but it also follows
H O that the bonds forming part of the second ring can be
HH H considered to share a trans or cis relationship to each
other. It is usual practice to show the stereochemistry
HO in the former way, via the ring fusion substituents.
diosgenin

108 STEREOCHEMISTRY

The situation is in many ways analogous to trans- and these afford useful comparisons as we consider
and cis-1,2-dimethylcyclohexane (see Section 3.3.2), conformational changes.

H ax cannot achieve
eq bonding within a
H six-membered ring
eq H
ax
H
trans-decalin ax

relate to eq
eq

ax

Now, trans-decalin forms a rather rigid system, and it change would require these to become axial. It is
transpires that the only conformational mobility possible impossible to join the two axial bonds into a ring
is ring flip of chairs to very much less favourable boats. system as small as six carbons; hence, there is no
Since both bonds of the second ring are equatorial with conformational mobility.
respect to the first ring, any other type of conformational

ax H ax rotate 60˚ eq ax
H eq H
eq H ≡
H H
cis-decalin
ax
ax eq

relate to ax
rotate 60˚
eq
≡

eq

On the other hand, cis-decalin is conformationally Since the second ring in trans-decalin effectively
mobile, and a simultaneous flipping in both rings introduces two equatorial substituents to the first ring,
produces a new conformer of equal energy. This is whilst in cis-decalin it provides one equatorial and one
not easy to visualize. In the scheme, the middle axial substituent, it is logical to predict that trans-
conformer has one ring viewed face on, so that we decalin should have a lower energy than cis-decalin.
have resorted to rotation of the structure to get an This is indeed the case, the energy difference being
appreciation of the new conformer with its rings in about 12 kJ mol−1.
chair form. It is best to have models to appreciate this
conformational flexibility. It is quite clear, though, that When we considered trans- and cis-1,2-dimethyl-
an axial bond becomes equatorial and an equatorial cyclohexane, we found that only three configurational
one becomes axial, just as with substituents in the cis- isomers exist, enantiomeric forms of the trans isomer,
1,2-dimethylcyclohexane analogue (see Section 3.3.2). together with the cis isomer, which is an optically
However, it is probably reassuring to appreciate that this inactive meso compound (see Section 3.4.5). The meso
conformational flexibility in two cis-fused cyclohexane relationship could be deduced from the plane of
rings is lost when a third ring is fused on, and in many symmetry in the hexagon representation.
of the fused ring systems of interest to us it becomes of
no further consequence.

CONFIGURATIONAL ISOMERS 109

no plane of destroy symmetry
symmetry
trans
trans two enantiomers
two enantiomers

plane of symmetry

cis three configurational isomers cis
meso two enantiomers
four configurational isomers
When we look at the structures of trans- and cis-
decalin, it is apparent that a further plane of symmetry, dimethyl substitution removes
through the ring fusion, is present in both structures. symmetry without adding a
This means that each isomer is superimposable on new chiral centre
its mirror image; consequently, there are only two
configurational isomers of decalin, one trans and HH
one cis.
HH
plane of symmetry mirror image H trans
H H
two enantiomers
≡≡
HH
H HH
HH
trans-decalin H cis
H
planes of symmetry two enantiomers
mirror image
four configurational isomers
HH
Fusing rings of different sizes can produce significant
≡≡ restraints, especially when rings of less than six carbons
are involved. However, the characteristics of these fused
HH systems can be deduced logically by applying our
knowledge of single ring systems.
cis-decalin
Fusion of a five-membered ring to a six-membered
only two configurational isomers ring gives a hydrindane system, and, as with decalins,
cis and trans forms are possible. Because the cyclopen-
The situation in trans- and cis-decalin is compli- tane ring is more planar than a cyclohexane ring (see
cated by the symmetry elements. If this symmetry is Section 3.3.2), this causes deformation and increases
destroyed, e.g. by introducing dimethyl substituents, strain at the ring fusion. This deformation is more eas-
we get back to reassuringly familiar territory in which ily accommodated with the cis-fusion than the trans-
two chiral centres lead to four configurational iso- fusion, and, in contrast to the decalins, the cis isomer
mers. The same is true in the trans- and cis-1,2- has a lower energy than the trans isomer (by about
dimethylcyclohexane series. 1 kJ mol−1). As in the decalins though, the cis form
is conformationally mobile, whereas the trans form
is fixed.

110 STEREOCHEMISTRY

H H
eq
H
trans-hydrindane eq
H

H ax H ax rotate 60˚ eq ax
H eq H
H eq H ≡
cis-hydrindane H H

The fusion of rings of different sizes reduces anticipated three isomeric forms, two enantiomeric
symmetry in the structures; instead of the rather unusual trans isomers and a meso cis isomer (compare 1,2-
situation with the decalins, where there are only two dimethylcyclohexane).
configurational isomers, the hydrindanes exist in the

HH plane of symmetry H H
H
HH ≡
trans
H
two enantiomers cis
meso

Box 3.18
Isomerizations influenced by ring fusions

Epimerization of cis-decalone If cis-decalone is treated with mild base, it is predominantly isomerized

to trans-decalone. This can be rationalized by considering stereodrawings of the two isomers.

HH
base

H H
O O

cis-decalone trans-decalone

H OH H H
eq HO H eq
H
ax eq
O
O
OH
base removes acidic proton
α to carbonyl and generates reformation of keto form;
enolate anion proton is acquired on lower face
to produce more stable isomer

The ring fusion in cis-decalone means that bonds forming the second ring have a relationship to the first ring
in which one bond is equatorial and one axial. In contrast, both such bonds in trans-decalone are equatorial to the

CONFIGURATIONAL ISOMERS 111

first ring. We can predict, therefore, that trans-decalone has a lower energy than cis-decalone. The isomerization
is brought about because the carbonyl group is adjacent (α) to the hydrogen at the ring fusion. This hydrogen is
relatively acidic and may be removed by base, generating the enolate anion (see Sections 4.3.5 and 10.1 for detail
of this reaction). The enolate anion must now be planar around the site of ring fusion and, by a reversal of the
process, may pick up a proton from either side of the double bond. However, instead of getting a 1 : 1 mixture of
the two possible isomers, this reaction very much favours the trans isomer because of its lower thermodynamic
energy. The equilibrium mixture contains principally trans-decalone.

Epimerization of etoposide The anticancer agent etoposide contains a five-membered lactone function

that is significantly strained because it is trans-fused. This material is readily converted into a relatively strain-
free cis-fused system by treating with very mild alkali, e.g. traces of detergent, and produces an epimer (see
Section 3.4.4) called picroetoposide. This isomer has no significant biological activity.

O O base removes acidic proton reformation of keto form OR
O α to carbonyl and results in change of
O generates enolate anion stereochemistry O
HO OH OR HO

O O O
O
O O O
HB O
NaOAc O

HO

B

MeO OMe MeO OMe MeO OMe
OH
OH OH

etoposide picroetoposide

The epimerization can be formulated as involving an enolate anion, as above (see Section 10.8). However, in
contrast to the decalin example above, cis-hydrindane is of lower energy than trans-hydrindane. In this particular
case, on reverting back to a carbonyl compound, the planar enolate anion is presented with the alternatives of
receiving a proton from one face to form a strained trans-fused system, or from the other face to form a strain-free
cis-fused system. The latter is very much preferred, so much so that the conversion of etoposide into its epimer is
almost quantitative. Although we can rationalize this behaviour simply by considering the hydrindane-type rings,
the fusion of this system to an aromatic ring causes additional distortion (see below), and the effect becomes
even more pronounced in favour of the cis-fused system.

This behaviour contrasts with the racemization of hyoscyamine to atropine, which also involves an enolate
anion derived from an ester system (see Section 10.8). As the term racemization implies, atropine is a 50 : 50
mixture of the two enantiomers. It shows how the proportion of each epimer formed can be influenced by other
stereochemical factors.

The fusion of a three-membered ring onto a six- can, therefore, only be cis-fused, and the six-membered
membered ring has much more serious limitations. A ring is forced to adopt the half-chair conformation we
three-membered ring must be planar, so it will distort saw with cyclohexene (see Section 3.3.2). There will be
the ring it is being fused to, and this restricts stereo- conformational mobility in this ring provided that there
chemical possibilities. For example, epoxycyclohexane are no other ring fusions to prevent this.

H H OO
O
O≡ H HH H
H
H 6-membered ring adopts
half-chair conformation
epoxycyclohexane

112 STEREOCHEMISTRY

Note that, in situations where a ring fusion produces Note that a cyclohexane system will be forced
chiral centres, we can find the number of configurational into a similar half-chair conformation by fusing a
isomers possible is less than that predicted from the 2n planar aromatic ring onto a cyclohexane ring (a
guidelines. This may be the consequence of symmetry, tetrahydronaphthalene system).
in that an isomer is the same as its mirror image, as we
have seen above. However, it can also be the result of tetrahydronaphthalene
restrictions caused by the ring fusion, so that one centre
effectively defines the chirality of another, thus reducing
the number of combinations. In epoxycyclohexanes, no
trans-fused variants can exist.

HH cyclohexene ring adopts half-
dimethyl substitution chair conformation

O O removes symmetry
without adding a new

H H chiral centre

two chiral centres, but only
two configurational isomers;
no trans isomers can exist

Box 3.19

Shapes of steroids

Steroids all contain a tetracyclic ring system comprised of three six-membered rings and one five-membered ring
fused together. Cholesterol is the best known of the steroids. It is an essential structural component of animal
cells, though the presence of excess cholesterol in the blood is definitely associated with the incidence of heart
disease and heart attacks.

Whilst cholesterol typifies the fundamental structure, further modifications to the side-chain and the ring system
help to create a wide range of biologically important natural products, e.g. sterols, steroidal saponins, cardioactive
glycosides, bile acids, corticosteroids, and mammalian sex hormones. Because of the profound biological activities
encountered, many natural steroids, together with a considerable number of synthetic and semi-synthetic steroidal
compounds, are routinely employed in medicine. The markedly different biological activities observed emanating
from compounds containing a common structural skeleton is, in part, ascribed to the functional groups attached
to the steroid nucleus and, in part, to the overall shape conferred on this nucleus by the stereochemistry of ring
fusions.

Let us start with cholestane, which is the basic hydrocarbon skeleton of cholesterol. This structure has all
ring fusions trans, and by logical extension of trans-decalin and trans-hydrindane can be deduced to have
approximately the shape illustrated. Because of the trans fusions, there is no conformational mobility except for
the unlikely flipping of ring A into a boat form, which we can ignore. The overall shape of cholestane is a rather
rigid and flattish structure. The rings are designated A–D as indicated.

HC D H H H
CD
A BH H AB H
H
H
cholestane H
all-trans

CONFIGURATIONAL ISOMERS 113

Cholesterol has a double bond in ring B at the A–B ring fusion, so this distorts the rings by demanding
that the arrangement around the double bond is planar. It is not possible to depict this perfectly in a typical
two-dimensional representation.

HH

H CD
H
H

5 H H HO AB
H H
HO 6

cholesterol ∆5-unsaturation Note: ∆5 is a neat way of indicating
that there is a double bond at position 5

The natural progestogen hormone progesterone also has a double bond at the A–B ring fusion, but this time
in ring A, so a similar distortion in ring A is required.

O

HO

H CD

AB H

45 H H O HH

O

progesterone ∆4-unsaturation

The fungal sterol ergosterol has double bonds at positions 5 and 7, both in the B ring, which consequently
should become essentially planar. The picture shown is a rough approximation. The antifungal effect of polyene
antibiotics, such as amphotericin and nystatin (see Box 7.14), depends upon their ability to bind strongly to
ergosterol in fungal membranes. They do not bind significantly to cholesterol in mammalian cells, so this provides
selective toxicity. The binding to ergosterol is very much influenced by the changes in shape conferred by the
extra double bond in ring B.

8 ABH H H
CD H
H
5H H HO
7 H
HO 6

ergosterol ∆5,7-unsaturation

In oestrogens, such as estradiol, the A ring is aromatic. Consequently, this ring is planar and distorts ring
B accordingly; again, it is difficult to draw this perfectly. The stereochemical outcome makes oestrogens seem
rather more flattened than the original all-trans arrangement in cholestane.

114 STEREOCHEMISTRY

Box 3.19 (continued)

OH

H OH

H HO A CD H

HH BH H
HO
A ring aromatic
estradiol

More dramatic changes are made to the shape of the steroid skeleton if ring fusions become cis rather than
trans. The most important examples involve the A–B and C–D ring fusions. It is not difficult to work out how
the modified skeleton looks after these changes. The approach is to start from the all-trans system and to delete
the appropriate ring, though retaining the bonds to the unchanged part as a guide to putting in the new ring. This
provides us with three of the bonds in the new ring, and it is just necessary to fill in the rest, using earlier decalin
or hydrindane templates.

cleave off appropriate ring, H
CD
leaving residual bonds
H

AB CD AB use residual bonds to
H H H form basis of new rings

H

all-trans

H H

H AB CD C D
H H H
AB
H

H

A–B cis C–D cis

The approach is used to show the shape of cholic acid, one of the bile acids secreted into the gut to emulsify
fats and encourage digestion. Cholic acid is characterized by a cis fusion of rings A and B.

HO CO2H HH H CO2H
CD
H AB H
H OH
5H H H H
HO OH H
H OH
H
cholic acid OH A–B cis

Digitoxigenin has cis fusions for both A–B and C–D rings. Glycosides of digitoxigenin are the powerful
heart drugs found in the foxglove, Digitalis purpurea. Note how a cis ring fusion changes the more-or-less flat
molecule of cholestane into a molecule with a significant ‘bend’ in its shape; digitoxigenin has two such ‘bends’.
These features are important in the binding of steroids to their receptors, and partially explain why we observe
quite different biological activities from compounds containing a common structural skeleton.

CONFIGURATIONAL ISOMERS 115

OO

O
O

H

H 14 AB CD H
OH
5 H OH H
HO HO H

H

digitoxigenin H A–B cis, C–D cis

Most natural steroids have the stereochemical features seen in cholesterol, though, as we have seen, there may
be some variations, particularly with respect to ring fusions affecting the A and D rings. Note that trans fusion at
the hydrindane C–D ring junction is energetically less favourable than a cis fusion (see Section 3.5.2), but most
natural steroid systems actually have this trans fusion.

HH

CD H

AB H
H
HH

∆5-unsaturation

O

H HO
HF
H O
HH H E
O

HO H
HH
HO
diosgenin H
diosgenin

We have met diosgenin as an example of a natural spiro compound (see Box 3.17), and further examination
of the structure shows the 5 double bond as in cholesterol, a second five-membered ring cis-fused onto the

five-membered ring D, as well as the spiro fusion of a six-membered ring. Before this structure dismays you,

take it slowly and logically. It should not be too difficult to end up with the stereodrawing shown here.

Box 3.20

The shape of penicillins

Penicillins are the most widely used of the clinical antibiotics. They contain in their structures an unusual fused
ring system in which a four-membered β-lactam ring is fused onto a five-membered thiazolidine. Both rings are
heterocyclic, and one of the ring fusion atoms is nitrogen. These heteroatoms do not alter our understanding of
molecular shape, since we can consider that they also have an essentially tetrahedral array of bonds or lone pair
electrons (see Section 2.6.3).

We have seen that, in cyclobutane and cyclopentane, a lower energy conformation is attained if the rings are not
planar (see Section 3.3.2). If one fuses a five-membered ring onto a four-membered ring, models demonstrate that
it is only possible to have a cis fusion in such a structure, and that conformational freedom in the four-membered
ring disappears if we are to achieve this bonding; the four-membered ring reverts to a more planar shape. It is
still possible to have the five-membered ring non-planar, thereby reducing eclipsed interactions.

116 STEREOCHEMISTRY

HH 3.5.3 Bridged ring systems
NS
In bridged ring compounds, rings share more than two
ON CO2H atoms, and the bridge can consist of one or more atoms.
O We have already met an example in bornane (see
Section 3.3.2), which we used as an illustration of how a
benzylpenicillin cyclohexane ring can be forced into a boat conformation
(penicillin G) to achieve the necessary bonding.

H2N H
O
S ≡
6 51
2
7 N4 3

bornane cyclohexane ring
forced into boat
CO2H conformation

6-aminopenicillanic acid

H2N SH If we inspect the ring system of bornane, omitting the
methyl groups, we can see that there are actually several
H H bridges of different lengths spanning the bridgehead
ON HO2C atoms, depending upon which atoms are considered.
This is used in nomenclature, as illustrated below,
can only be cis-fused including in square brackets all the bridges, listed in
four-membered ring is planar decreasing lengths. Numbering, when necessary, always
five-membered ring is non-planar starts from a bridgehead atom. A closer inspection of the
ring fusion forces N into one configuration shape of bicyclo[2,2,2]octane (best with a model), which
has two-carbon bridges, shows that each ring system has
The cis fusion in which one of the fusion atoms is the boat conformation.
nitrogen merely indicates that the nitrogen lone pair
electrons occupy the remaining part of the tetrahedral bridgehead one-atom bridge
array. It does, however, mean that inversion at the two-atom bridge
nitrogen atom (see Section 3.4.1) is not possible,
since that would hypothetically result in formation bridgehead
of the impossible trans-fused system. The ring
fusion has thus frozen the nitrogen atom into one two-atom bridge
configuration.
bicyclo[3,1,1]heptane bicyclo[2,2,1]heptane
Fusion of a four-membered ring onto a six-
membered ring is also only possible with a cis
fusion; cephalosporins provide excellent examples
of such compounds, and the comments made above
for penicillins are equally valid for these compounds.

HH
H2N N S

bicyclo[2,2,2]octane

CO2H ON O
O O
CO2H ≡

cephalosporin C

bicyclo[2,2,2]octane all rings boat

CONFIGURATIONAL ISOMERS 117

Note that the ring systems with small bridges plane of CO2H CO2H
illustrated here can have no conformational mobility, symmetry CO2H
and are quite fixed. Bornane also has no configurational
isomers. If we are going to bridge a cyclohexane ring H3C CO2H cis plane of
with a one-carbon bridge, there is only one way to H3C symmetry
achieve this; in other words, the configuration at the
second bridgehead is fixed by that chosen at the first. trans
A similar situation confronted us with fused rings,
in that, in order to achieve the fusion of a small When we move on to camphor, a ketone deriva-
ring, only a cis fusion was feasible (see Section 3.5.2). tive of bornane, we find this can exist in two enan-
Furthermore, bornane has a plane of symmetry and tiomeric forms because the plane of symmetry has been
can be superimposed on its mirror image, so only one destroyed. Nevertheless, there are only two configura-
configurational isomer can exist. tional isomers despite the presence of two chiral centres;
bridging does not allow the other two variants to exist.
mirror image

≡≡

bornane OO

plane of (−)-camphor OO (+)-camphor
symmetry
two chiral centres
only two stereoisomers

this type of bridging β-Pinene is representative of a bicyclo[3,1,1]heptane
is stereochemically system. This natural product has two chiral centres, but
impossible can exist only in the (+)- and (−)-enantiomeric forms
shown.
We should compare this system with a 1,4-
disubstituted cyclohexane such as 4-methylcyclo- (−)-β-pinene two chiral centres (+)-β-pinene
hexanecarboxylic acid (see Section 3.4.4). There is a only two stereoisomers
plane of symmetry in this molecule, so there are no
chiral centres; but geometric isomers exist, allowing
cis and trans stereoisomers. The restrictions imposed
by bridging have now destroyed any possibility of
geometric isomerism.

Box 3.21

Stereochemistry of tropane alkaloids

The tropane alkaloids (−)-hyoscyamine and (−)-hyoscine are found in the toxic plants deadly nightshade (Atropa
belladonna) and thornapple (Datura stramonium) and are widely used in medicine. Hyoscyamine, usually in the
form of its racemate atropine, is used to dilate the pupil of the eye, and hyoscine is employed to control motion
sickness. Both alkaloids are esters of (−)-tropic acid.

The alcohol portion in hyoscyamine is tropine; in hyoscine it is the epoxide scopine. Tropine is an example
of an azabicyclo[3,2,1]octane system with a nitrogen bridge, whereas scopine is a tricylic system with a three-
membered epoxide ring fused onto tropine. Note that systematic nomenclature considers an all-carbon ring system
with one carbon replaced by nitrogen; hence, tropane is an azabicyclooctane (see Section 1.4).

There are several interesting stereochemical features accommodated within these structures. First, both tropine
and scopine are optically inactive meso compounds; despite the chiral centres, two for tropine and four for

118 STEREOCHEMISTRY

Box 3.21 (continued)

Me N * Me
* N*

H CH2OH O* * H CH2OH * chiral centres
O * O

* *

O O

(–)-hyoscyamine (–)-hyoscine
(scopolamine)
plane of Me plane of
Me N symmetry N symmetry

12 NMe OH ≡ O OH
tropine OH
7

8 NMe 3

6 4
5

tropane tropine scopine

N-methyl-8-azabicyclo[3,2,1]octane

scopine, both compounds have a plane of symmetry, so that optical activity conferred by one centre is cancelled
out by its mirror image centre. The optical activities of hyoscyamine and hyoscine are derived entirely from the
chiral centre in the tropic acid portion. Atropine, the racemic form of hyoscyamine, is the ester of tropine with
(±)-tropic acid (see Box 10.9).

Me N Me nitrogen inversion can occur:
N the methyl group is preferentially
OH equatorial in tropine but axial in scopine
tropine O (minimizes interaction with epoxide)
OH

scopine

Note also that, although we normally see rapid inversion at a nitrogen atom, the N-methyl group in hyoscyamine
is preferentially in the lower energy equatorial position of the chair-like piperidine ring, as would be predicted.
However, in hyoscine, the N-methyl group has been found to be axial, not the expected equatorial. This seems
to arise to minimize interaction with the extra epoxide ring in scopine.

* chiral centres CO2Me Me N CO2Me
* OH
Me N * *O
*
O

cocaine (−)-methylecgonine

When we look at another tropane alkaloid, cocaine, we get a different scenario. Cocaine is obtained from the
coca plant Erythroxylum coca, and is a powerful local anaesthetic, but now known primarily as a drug of abuse.
There is no chiral centre in the acid portion, which is benzoic acid, but the optical activity of cocaine comes from
the alcohol methylecgonine. Because of the ester function in methylecgonine, the tropane system is no longer
symmetrical, and the four chiral centres all contribute towards optical activity.

CONFIGURATIONAL ISOMERS 119

Me N R CO2Me Me N R Me N R
S R S S
S OH sH r OH
OH H
H

(−)-methylecgonine tropine pseudotropine

Now, you may have noticed that the hydroxyl group in methylecgonine is oriented differently from that in
tropine. In methylecgonine it is easy to define the position of the hydroxyl, since this is a chiral centre and we
can use the R/S nomenclature. An alternative stereoisomer of tropine exists, and this is called pseudotropine.
How can we define the configuration for the hydroxyl when the plane of symmetry of the molecule goes through
this centre and means this centre is not chiral but can exist in two different arrangements?

This is a situation allowed for in the IUPAC nomenclature rules, because if we are faced with two groups which
are the same but have opposite chiralities, then the group with R chirality has a higher priority than the group
with S chirality. Applying this rule, tropine would have the S configuration and pseudotropine the R configuration
at this centre. Because of the plane of symmetry, these atoms are not strictly chiral, and this is taken into account
by using lower-case letters; tropine is s and pseudotropine is r.



4

Acids and bases

4.1 Acid–base equilibria The Lewis definition of acids and bases is
rather more general than the Brønsted–Lowry version
A particularly important concept in chemistry is that (which refers to systems involving proton transfer) in
associated with proton loss and gain, i.e. acidity and that:
basicity. Acids produce positively charged hydrogen
ions H+ (protons) in aqueous solution; the more • an acid is an electron-pair acceptor;
acidic a compound is, the greater the concentration of
protons it produces. In water, protons do not have an • a base is an electron-pair donor.
independent existence, but become strongly attached
to a water molecule to give the stable hydronium Thus, Lewis acids include such species as boron
ion H3O+. In the Brønsted–Lowry definition: trifluoride, which is able to react with trimethylamine
to form a salt.
• an acid is a substance that will donate a proton;
Me3N BF3 Me3N BF3
• a base is a substance that will accept a proton.
Lewis Lewis
Thus, in water, the acid HCl ionizes to produce H3O+ base acid
and Cl− ions.

H2O H Cl H3O Cl There is no fundamental difference between
trimethylamine acting as a Brønsted base or as a
base acid conjugate acid conjugate Lewis base, except that in the Brønsted concept it
(proton (proton of H2O base of HCl donates its electrons to a proton electrophile, whereas
acceptor) donor) as a Lewis base it donates its electrons to a Lewis acid
electrophile.
H3O+ is termed the conjugate acid (of the base
H2O) and Cl− is termed the conjugate base (of the R3N H R3N H
acid HCl). In general terms, cleavage of the H–A
bond in an acid HA is brought about by a base, Brønsted conjugate
generating the conjugate acid of the base, together base acid
with the conjugate base of the acid. You may wish to
read that sentence again!

B HA BH A R3N E R3N E
base acid
conjugate conjugate Lewis
acid base base

Essentials of Organic Chemistry Paul M Dewick
 2006 John Wiley & Sons, Ltd

122 ACIDS AND BASES

4.2 Acidity and pKa values Or, put another way:

For the ionization of the acid HA in water • the smaller the value of pKa, the stronger is the
acid;
K
• the larger the value of pKa, the weaker is the
H2O + HA H3O + A acid.

the equilibrium constant K is given by the formula We find that pKa values range from about −12 to
52, but it must be appreciated right from the start that
K = [A−][H3O+] a difference of one pKa unit actually represents a 10-
[HA][H2O] fold difference in Ka and, thus, a 10-fold difference in
H3O+ concentration. A twofold difference in acidity
where [HA] signifies the concentration of HA, etc. would be indicated by a pKa difference of just 0.3
However, because the concentration of water is units (log 2 = 0.3). Accordingly, a difference of n
pKa units indicates a 10n-fold difference in acidity,
essentially constant in aqueous solution, a new so that the range −12 to 52 actually represents a
equilibrium constant Ka is defined as huge factor of 1064. A compound with pKa < 5
is regarded as a reasonably strong acid, and those
Ka = [A−][H3O+] with pKa < 0 are very strong acids. At first glance,
[HA] negative pKa values seem rather strange, but this
only means that the equilibrium lies heavily towards
Ka is termed the acidity constant, and its magnitude ionization; Ka is large and, therefore, pKa = − log Ka
allows us to classify acids as strong acids (a becomes negative.
large value for Ka and, consequently, a high H3O+
concentration) or weak acids (a small value for Ka H2O + HA H3O + A
and, thus, a low H3O+ concentration). For example,
the strong acid HCl has Ka = 107. However, for pKa = −log10 Ka
weak acids, the amount of ionization is much less

and, consequently, the value of Ka is rather small.
Thus, acetic acid CH3CO2H has Ka = 1.76 × 10−5.
To avoid using such small numbers as these, Ka is
usually expressed in the logarithmic form pKa where

pKa = − log10 Ka Ka = 0.01 Ka = 0.1 Ka = 1 Ka = 10 Ka = 100

Accordingly, the pKa for acetic acid is 4.75: increasing acid strength
pKa = − log(1.76 × 10−5) = −(−4.75) = 4.75
pKa = 2 pKa = 1 pKa = 0 pKa = −1 pKa = −2
The pKa for hydrochloric acid can similarly be
calculated to be −7: As we use pKa values, we shall find that, in most
cases, relative, rather than specific, values are all
pKa = − log(107) = −7 we need to consider to help us predict chemical
behaviour and reactivity. Thus, from pKa values, we
This means there is an inverse relationship between can see that acetic acid (pKa 4.75) is a weaker acid
the strength of an acid and pKa: than hydrochloric acid (pKa − 7).

• a strong acid has a large Ka and, thus, a small pKa, pKa values for a wide variety of different com-
i.e. A− is favoured over HA; pounds are given in Tables 4.1–4.6. Compounds are
listed in order of increasing acidity. Although pKa
• a weak acid has a small Ka and, thus, a large pKa, values included extend from about 52 to −10, values
i.e. HA is favoured over A−. in the middle of the range are known most accurately.
This is because they can be measured readily in aque-
ous solution. Outside of the range from about 2 to 12,
pKa values have to be determined in other solvents,

ACIDITY AND pKa VALUES 123

Table 4.1 pKa values of H–X acids Table 4.3 pKa values of N–H, O–H, and S–H acids

Acid Conjugate base pKa Acid Conjugate base pKa

CH4 CH3 48 NH3 NH2 38
NH3 NH2 38 36
H2 H 35 (CH3)2CH (CH3)2CH
H2O HO 15.7 NH N 35
H–C≡N 28
H2S C≡N 9.1 (CH3)2CH (CH3)2CH 15
HF HS 7
F 3.2 CH3CH2NH2 CH3CH2NH
H3PO4 H2 PO4 2.1 Ph–NH2 Ph–NH
HNO3 NO3 −1.4 CH3CONH2 CH3CONH
H2SO4 HSO4 −3.0
HCl Cl −7 (CH3)3C–OH (CH3)3C–O 19
HBr Br −9 CH3OH CH3O 15.5
HI I −10 CH3CH2OH CH3CH2O 16
H2O HO 15.7
Ph–OH Ph–O 10

Table 4.2 pKa values of C–H acids CH3SH CH3S 10.5
Conjugate base HS 7
Acid pKa H2S Ph–S 6.5
Ph–SH

H 52

H H Table 4.4 pKa values of CO2H and SO3H acids
50
H3C–CH3 H3C–CH2
CH4 CH3 48 Acid Conjugate base pKa
H2=CH2 H2=CH
44 CH3CO2H CH3CO2 4.8
H 44 Ph–CO2H Ph–CO2 4.2

H2C=CH–CH3 HCO2H HCO2 3.7

CH3 H2C=CH–CH2 43 ClCH2CO2H ClCH2CO2 2.9
Cl2CHCO2H Cl2CHCO2 1.3
Ph3C–H CH2 41 Cl3CCO2H Cl3CCO2 0.7
H3C–C≡N F3CCO2 −0.3
HC≡C–H Ph3C F3CCO2H
H2C–C≡N
32 Me–SO3 −1.2
HC≡C 25 Me–SO3H SO3H H3C −1.3
25 H3C SO3

or even by indirect methods; results are then extrapo- However, the fact that they do have to be measured
means that as you look in the literature for the pKa of
lated to give the value in water. The figures presented a particular compound you may find slightly different
values can be presented. Do not let this confuse you.
in Tables 4.1–4.6 have been intentionally rounded to As mentioned above, relative, rather than specific,
values are our main concern.
stress that a high level of accuracy is usually inap-
We have chosen to present the pKa values as a
propriate. series of tables, rather than in a single one. This
The range of pKa values that can be measured in should help you to locate a particular compound

water is determined by the ionization of water itself,
i.e. −1.74 (the pKa of H3O+) to 15.74 (the pKa
of H2O); see Box 4.1. Acids that are stronger than
H3O+ simply protonate water, whereas bases that
are stronger than HO− remove protons from water.

124 ACIDS AND BASES

Table 4.5 pKa values of CH–CO, CH–CN, and CH–NO2 acids Table 4.6 (continued)

Acid Conjugate base pKa

CH3CO2CH2 CH3CO2CH3 24 H3O H2O −1.7
CH3COCH3 CH3COCH2 19 PhCONH2 −2.2
CH3CH=O 17 OH
CH2CH=O 13 Ph C CH3OH −2.2
CH3O2C −2.4
CH2 CH3O2C NH2 CH3CH2OH −3.8
CH −3.8
CH3O2C CH3OH2 (CH3)3C–OH
CH3O2C H3C −6.1
CH3CH2OH2
CH3COCH2CO2CH3 CH3COCHCO2CH3 11 O −6.5
(CH3)3C–OH2 H3C
CH3COCH2COCH3 CH3COCHCOCH3 9 −6.7
H3C O −7
H3C–C≡N H2C–C≡N 25 OH H3C C −7.2
CH3NO2 CH2NO2 10
H3C OH
Table 4.6 pKa values of N+, O+, and S+ acids O
OH H3C C
Acid Conjugate base pKa H3C C OCH3

H2N H2N 13.6 OH
C NH2 C NH
OH
H2N H2N H3C C

CH3NH3 CH3NH2 10.6 OCH3

(CH3)2NH2 (CH3)2NH 10.7 Ph–OH2 Ph–OH
PhCH=OH PhCH=O
(CH3)3NH (CH3)3N 9.8 (CH3)2C=OH (CH3)2C=O
NH4 NH3
9.2
5.2

N N CH3CH=OH CH3CH=O −8
H −5.4
Ph–N(CH3)2 H3C H3C
Ph–NH(CH3)2 Ph–NH2 5.1 S H S −6.8
Ph2NH
Ph–NH3 H3C–C≡N H3C H3C
CH3CONH2 4.6
Ph2NH2 CH3SH
CH3SH2
H3C–C≡NH 0.8
OH
−10 according to its functional group, and we hope that
H3C C −1.4 this will also emphasize similarities and differences
NH2 in related structures. It also means that you may find
(continues ) some examples turning up in more than one table.

As we consider different aspects of chemical reac-
tivity in subsequent chapters, we shall see how pKa

ELECTRONIC AND STRUCTURAL FEATURES THAT INFLUENCE ACIDITY 125

values can be used to predict whether a reagent is a 4.3.3 Inductive effects
good or a poor nucleophile, whether it can function as
a good leaving group, and how easy it is to generate Electron-donating and electron-withdrawing groups
anionic nucleophiles. We shall also find that pKa influence acidity by respectively destabilizing or
values can tell us how much of a compound or a drug stabilizing the conjugate base. This inductive effect,
is ionized under particular conditions and, therefore, a charge polarization transmitted through σ bonds
whether or not it can be produced in a soluble form. (see Section 2.7), causes a shift in electron density,
and its influence may easily be predicted.
It is now appropriate to consider some of the
electronic and structural features that influence pKa XAH XA H
so that we can rationalize and predict relative
acidities.

4.3 Electronic and structural features XA XA
that influence acidity
electron-withdrawing electron-donating
4.3.1 Electronegativity inductive effect inductive effect
stabilizing
The more electronegative an element is, the more it destabilizing
helps to stabilize the negative charge of the conjugate
base. For example, the acidities of compounds of Thus, electron-withdrawing groups increase
second-row elements in the periodic table increase acidity:
as the atom to which hydrogen is attached becomes
more electronegative: • pKa values for the simple carboxylic acid acetic
acid and its halogenated derivatives chloroacetic
• pKa values for CH4, NH3, H2O and HF are acid, dichloroacetic acid, and trichloroacetic acid
about 48, 38, 16 and 3, respectively, i.e. we are about 4.8, 2.9, 1.3, and 0.7 respectively, the
have increasing acidity left to right as the elec- inductive effects of the chlorine atoms spreading
tronegativity of the atom attached to hydrogen the charge of the conjugate base and thus stabiliz-
increases. ing it.

4.3.2 Bond energies O O
Cl
Within a single column of the periodic table, acidities OH
increase as one descends the column: pKa values for acetic acid OH
HF, HCl, HBr, and HI are about 3, −7, −9, and
−10 respectively, i.e. we have increasing acidity on pKa 4.8 chloroacetic acid
descending the group. pKa 2.9

This is the reverse of what might be expected O O
simply based on electronegativities, but relates to the Cl
increasing size of the atom and the corresponding Cl OH
improved ability to disperse the negative charge OH Cl
over the atom. We are seeing a weakening in bond Cl
strengths on descending the group. dichloroacetic acid Cl
pKa 1.3
Similarly, although sulfur is less electronega- trichloroacetic acid
tive than oxygen, thiols (RSH) are more acidic
than alcohols (ROH). For example, pKa values for pKa 0.7
methanethiol and methanol are 10.5 and 16 respec-
tively. acidity increases as the number of

electron-withdrawing substituents

increases

O

Cl

Cl O
Cl

126 ACIDS AND BASES

• Increasing the number of halogen atoms increases O I O O
this effect, with a consequent increase in acidity. OH OH Br
Note that introduction of one chlorine atom
increases acidity by a factor of almost 100, and OH
trichloroacetic acid is a strong acid.
acetic acid iodoacetic acid bromoacetic acid
• Because of the different electronegativities of pKa 4.8
the various halogens, we can also predict that pKa 3.2 pKa 2.9
fluorine will have a greater effect than chlorine,
which in turn will increase acidity more than OO
bromine or iodine. This is reflected in the observed Cl F
acidities of monohalogenated acetic acids, though
the increased acidity of chloroacetic acid (pKa OH OH
2.87) over bromoacetic acid (pKa 2.90) is not
apparent because of the rounding-up process. chloroacetic acid fluoroacetic acid

• The inductive effect is a rather short-range pKa 2.9 pKa 2.6
effect, and its influence decreases rapidly as the
acidity increases as
substituent becomes
more electronegative

O O Cl O Cl O
OH OH OH
OH
butanoic acid Cl 3-chlorobutanoic acid 4-chlorobutanoic acid
pKa 4.9 2-chlorobutanoic acid
pKa 2.9 pKa 4.1 pKa 4.5

effect of electronegative
substituent decreases as
it is located further away
from acidic group

Table 4.7 Inductive effects from functional groups

Electron-withdrawing groups Electron-donating groups

––F ––CO2H ––O
––Cl ––CO2R ––CH3

––Br O ––CO2
––I C
––OR N
––OH ––C≡N S
––NO2
N ––SO2––

––SR

––SH

ELECTRONIC AND STRUCTURAL FEATURES THAT INFLUENCE ACIDITY 127

substituent in question is located further away OO
from the site of negative charge, because it has
to be transmitted through more bonds. Thus, the H OH OH
effect on the acidity in butanoic acid deriva- formic acid acetic acid
tives can be seen to diminish with distance. 2-
Chlorobutanoic acid (pKa 2.9) shows a significant pKa 3.7 pKa 4.8
enhancement in acidity over butanoic acid (pKa
4.9), whereas 3-chlorobutanoic acid (pKa 4.1) and O O
4-chlorobutanoic acid (pKa 4.5) show rather more
modest changes. OH OH

• Other electron-withdrawing groups that increase propionic acid butanoic acid
the acidity of acids include, listed in decreasing pKa 4.9 pKa 4.8
order of their effect: –NO2, –N+R3, –CN, –CO2R,
–CO–, –OR and –OH. A more extensive list is electron-donating effect of alkyl
given in Table 4.7. groups is most marked on going
from formic acid to acetic acid
Electron-donating groups will have the opposite
effect, destabilizing the conjugate base by increasing O
electron density, and thus produce weaker acids. The
most common electron-donating groups encountered H3C O
are going to be alkyl groups, though the effect
from alkyl groups is actually rather small. Indeed, primarily related to solvation effects. In solution,
it is not immediately apparent why there should the conjugate base anion is surrounded with polar
be any inductive effect at all, since substitution of solvent molecules. This solvation helps to stabilize
hydrogen by alkyl should not lead to any bond the conjugate base, and thus increases the acidity of
polarization. At this point, we should merely note the alcohol. As we get more alkyl groups, solvation
that alkyl groups have a weak electron-donating of the anion is diminished because of the increased
effect – it may not be strictly an inductive effect (see steric hindrance they cause, and observed acidity
Section 6.2.1). also decreases.

• The pKa value for formic acid (pKa 3.7) makes H3C OH OH
it more acidic than acetic acid (pKa 4.8). The
electron-donating effect of the methyl group is methanol ethanol
most marked on going from formic acid to acetic pKa 15.5 pKa 16.0
acid, since the acidity of propionic acid (pKa 4.9)
and butanoic acid (pKa 4.8) vary little from that of OH OH
acetic acid. The electron-donating effect from alkyl
substituents is relatively small, being considerably isopropanol tert-butanol
smaller than inductive effects from most electron- pKa 17 pKa 19
withdrawing groups, and also rapidly diminishes
along a carbon chain. H O H3C O
C C
• Alcohols are much less acidic than carboxylic
acids; but, as one progresses through the sequence H H3C
methanol, ethanol, isopropanol, and tert-butanol, H H3C
pKa values gradually increase from 15.5 to 19, a
substantial decrease in acidity. Although this was alkyl groups hinder
originally thought to be caused by the inductive approach of solvation
effects of methyl groups, it is now known to be molecules

128 ACIDS AND BASES

4.3.4 Hybridization effects is sp2 (33% s character), and in acetylene it is sp
(50% s character). This makes alkynes (acetylenes)
The acidity of a C–H bond is influenced by the relatively acidic for hydrocarbons. It is also a
hybridization state of the carbon atom attached to the contributing factor in the acidity of HCN (pKa
acidic hydrogen. Dissociation of the acid generates 9.1), where the conjugate base cyanide is an sp-
an anion whose lone pair of electrons is held in a hybridized anion, though additional stabilization
hybridized orbital. We can consider sp orbitals to comes from the electronegative nitrogen atom.
have more s character than sp2 orbitals, and similarly
sp2 orbitals to have more s character than sp3 orbitals So far we have considered the hybridization state
(see Section 2.6.2). Since s orbitals are closer to of the orbital associated with the anionic charge.
the nucleus than p orbitals, it follows that electrons However, the hybridization state elsewhere in the
in an sp-hybridized orbital are held closer to the molecule may also affect acidity. The more s char-
nucleus than those in an sp2 orbital; those in an acter an orbital has, the closer the electrons are held
sp2 orbital are similarly closer to the nucleus than to the nucleus, and this effectively makes the atom
those in an sp3 orbital. It is more favourable for the more electronegative. This may be explained in terms
electrons to be held close to the positively charged of hybridization modifying inductive effects, such
nucleus, and thus an sp-hybridized anion is more that sp-hybridized carbons are effectively more elec-
stable than an sp2-hydridized anion, which is more tronegative than sp2-hybridized carbons, and simi-
stable than an sp3-hybridized anion. Thus, the acidity larly, sp2-hybridized carbons are more electronega-
of a C–H bond decreases as the s character of the tive than sp3-hybridized carbons.
bond decreases.
• The pKa values for the following acids illustrate
• The pKa of the hydrocarbon ethane is about 50, that, as the carbon atom adjacent to the carboxylic
that of ethylene about 44, and that of acetylene acid group changes from sp3 to sp2 to sp
is about 25. The hybridization of the C–H bond hybridization, the acidity increases, in accord with
in ethane is sp3 (25% s character), in ethylene it the electronegativity explanation above. Note that
benzoic acid (sp2 hybridization) has a similar pKa
H HH H HH to acrylic acid (propenoic acid), which also has sp2
H hybridization.

H H H O O
H sp3 orbital OH sp2

ethane HH sp3 OH
H propionic acid
pKa 50 acrylic acid
sp2 orbital pKa 4.9 (propenoic acid)
HH
H pKa 4.2
HH sp orbital
O O
ethylene NC sp2 sp
pKa 44 sp orbital
OH OH
HH
acetylene benzoic acid propiolic acid
pKa 25 pKa 4.2 (propynoic acid)

pKa 1.8

NCH inductive effect resulting
from hybridization
hydrocyanic acid
pKa 9.1

ELECTRONIC AND STRUCTURAL FEATURES THAT INFLUENCE ACIDITY 129

4.3.5 Resonance/delocalization effects However, drawing resonance structures provides a
simple and convenient way of predicting stability
Delocalization of charge in the conjugate base anion through delocalization (see Section 2.10).
through resonance is a stabilizing factor and will
be reflected by an increase in acidity. Drawing The pKa of ethanol is 16, and that of acetic acid
resonance structures allows us to rationalize that is 4.8. The increased acidity of acetic acid relative to
the negative charge is not permanently localized on ethanol can be rationalized in terms of delocalization
a particular atom, but may be dispersed to other of charge in the acetate anion, whereas in ethoxide
areas of the structure. We should appreciate that a the charge is localized on oxygen. Even more
better interpretation is that the electrons are contained delocalization is possible in the methanesulfonate
in a molecular orbital that spans several atoms. anion, and this is reflected in the increased acidity
of methanesulfonic acid (pKa − 1.2).

O O OO

OH O OH O OO
acetic acid
ethanol ethoxide anion acetate anion delocalization of charge
pKa 16 localized charge pKa 4.8 delocalized charge is sometimes depicted
via partial bonds

O O O O O

S OH SO SO SO SO
O O O O O

methanesulfonic acid methanesulfonate anion delocalization depicted
pKa − 1.2 delocalized charge via partial bonds

We have also shown some representations of The alkane propane has pKa 50, yet the presence of
acetate and methanesulfonate anions that have been the double bond in propene means the methyl protons
devised to emphasize resonance delocalization; these
include partial bonds rather than double/single in this alkene have pKa 43; this value is similar to
bonds. Although these representations are valu- that of ethylene (pKa 44), where increased acidity
able, they can lead to some confusion in inter- was rationalized through sp2 hybridization effects.
pretation. It is important to remember that there
is a double bond in these systems. Therefore, 1,3-Pentadiene is yet more acidic, having pKa 33 for
we prefer to draw out the contributing resonance the methyl protons. In each case, increased acidity
structures.
in the unsaturated compounds may be ascribed to

delocalization of charge in the conjugate base. Note

that we use the term allyl for the propenyl group.

H

propane
pKa 50

H resonance stabilized delocalization depicted
allyl anion via partial bonds
propene
pKa 43

H

1,3-pentadiene resonance stabilized delocalization depicted
pKa 33 pentadienyl anion via partial bonds

130 ACIDS AND BASES

Resonance stabilization is also responsible for the the charge to be transferred to the electronegative
increased acidity of a C–H group situated adjacent oxygen atom. As a result, acetone (pKa 19) is
to a carbonyl group. The anion is stabilized through significantly more acidic than propene (pKa 43).
delocalization of charge, similar to that seen with the Anions of this type, termed enolate anions, are some
allyl anion derived from propene; but this system is of the most important reactive species used in organic
even more favourable, in that delocalization allows chemistry (see Chapter 10).

O OO O
H H
H H H
H
H H H favoured − charge on delocalization depicted
acetone resonance-stabilized electronegative oxygen via partial bonds
pKa 19 enolate anion

The acidity of a C–H is further enhanced if it whereas the terminal protons adjacent to just a sin-
is adjacent to two carbonyl groups, as in the 1,3- gle carbonyl have pKa 19, similar to acetone above.
diketone acetylacetone. The enolate anion is stabi- It is clear that increased delocalization has a pro-
lized by delocalization, and both carbonyl oxygens found effect on the acidity. These two values should
can participate in the process. This is reflected in the be compared with that of the hydrocarbon propane
pKa 9 for the protons between the two carbonyls, (pKa 50).

OO OO OO OO OO
H
H H
resonance-stabilized
HH H HH
H enolate anion
pKa 19 delocalization depicted
pKa 9 via partial bonds

acetylacetone

Aromatic rings are themselves excellent examples ring and the carboxyl. The pKa of phenylacetic acid
(pKa 4.3), compared with acetic acid (pKa 4.8),
of resonance and delocalization of electrons (see demonstrates the inductive effect of a benzene ring.
However, we might then expect benzoic acid to
Section 2.10). They also influence the acidity of be a rather stronger acid than it actually is, since
the phenyl group is closer to the carboxyl group
appropriate substituent groups, as seen in benzoic than in phenylacetic acid. We attribute the lower
acid strength to an additional resonance effect in
acids. Benzoic acid (pKa 4.2) is a stronger acid the carboxylic acid that is not favourable in the
than acetic acid (pKa 4.8), and it is also stronger anion, where it would lead to a carboxylate carrying
than its saturated analogue cyclohexanecarboxylic a double negative charge; therefore, the resonance
effect weakens the acid strength.
acid (pKa 4.9). The phenyl group exerts an electron-
withdrawing effect because the hybridization of the
ring carbons is sp2; consequently, electrons are held
closer to the carbon atom than in an sp3-hybridized

orbital. This polarizes the bond between the aromatic

CO2H CO2H CO2H inductive effect CO2H
CH3 sp3 sp2 resulting from
acetic acid hybridization
pKa 4.8
cyclohexane-
carboxylic acid benzoic acid phenylacetic acid
pKa 4.2 pKa 4.3
pKa 4.9

ELECTRONIC AND STRUCTURAL FEATURES THAT INFLUENCE ACIDITY 131

O OH O OH O OH OO

resonance favours non-ionized benzoic acid resonance unfavourable
in anion

Further inductive effects from other substituents closer to the carboxyl group, and it will help to
enhance or counter these effects with predictable stabilize the conjugate base. The acid-weakening
results. Thus, a halogen such as chlorine, with resonance effects are also diminished by the inductive
a strong inductive effect, produces stronger acids, effects of halogens; it is not favourable to have an
especially in the case of the ortho derivative. electron-withdrawing substituent close to a positive
Here, the extra inductive effect is correspondingly charge.

CO2H CO2H CO2H O OH O OH
Cl

Cl Cl Cl Cl
pKa 4.0 pKa 3.8 unfavourable
CO2H pKa 2.9 substituent OH
CO2H O OH
CO2H inductive effects
CH3 O

CH3 CH3 pKa 3.9 CH3 CH3
pKa 4.4 pKa 4.3 favourable

On the other hand, methyl substituents have a position can have an influence on the carboxyl group,
weak electron-donating effect opposing that of the forcing it out of the plane of the ring. The result is
aromatic ring. This also favours resonance in the non- that resonance is now inhibited because the orbitals
ionized acid. There is only a modest effect on acidity, of the carbonyl group are no longer coplanar with
except when the methyl is in the ortho position, the benzene ring. In almost all cases, the ortho-
where the effect is closer to the carboxyl group. substituted benzoic acid tends to be the strongest acid
However, ortho substituents add a further dimension of the three isomers.
that is predominantly steric. Large groups in the ortho

O OH O OH O OH
X
steric hindrance distorts
carboxyl from coplanarity
with benzene ring and
inhibits resonance

resonance requires O
coplanarity of carbonyl
with benzene ring OH
X
e.g. rotation of carboxyl group;
carbonyl now at right angles to benzene
ring and orbitals cannot overlap

132 ACIDS AND BASES

When substituents can also be involved in the O OH O OH
resonance effects, changes in acidity become more
marked. Consider hydroxy- and methoxy-benzoic HO HO
acid derivatives. The pKa values are found to be
3.0, 4.1, and 4.6 for the ortho, meta, and para resonance stabilizes
hydroxy derivatives respectively, and 4.1, 4.1, and the non-ionized acid
4.5 respectively for the corresponding methoxy
derivatives.

(pKa values for CO2H group)

CO2H CO2H CO2H OO OO
OH

OH OH HO HO
pKa (CO2H) 4.6 pKa (CO2H) 4.1 pKa (CO2H) 3.0 resonance destabilizes the
conjugate base
CO2H CO2H CO2H
OCH3 O OH

OCH3

OCH3 pKa 4.1 pKa 4.1 OH
pKa 4.5
resonance delocalizes
Let us ignore the figure for ortho-hydroxybenzoic electrons only to ring carbons
acid for the moment, since there is yet another
feature affecting acidity. We then see that the para resonance structure (see Section 2.10). We shall use
derivatives are rather less acidic than we might ‘resonance effect’ rather than ‘mesomeric effect’ to
predict merely from the inductive effect of the OH avoid having the alternative terminologies.
or OMe groups. In fact, pKa values show that
these compounds are less acidic than benzoic acid, We can write a similar delocalization picture for
whereas the inductive effect would suggest they the ortho-substituted compounds, but this is coun-
should be more acidic. This is because of a large tered by the opposing inductive effect close to the car-
resonance effect emanating from the substituent in boxyl. However, the steric effect, as described above,
which electronic charge is transmitted through the means large groups in the ortho position can force
conjugated system of the aromatic ring into the the carboxyl group out of the plane of the ring. This
carboxyl group. weakens the resonance effect, since delocalization is
dependent upon coplanarity in the conjugate system.
The electron-donating effect originates from the
lone pair electrons on oxygen, with overlap into the π Resonance stabilization is not as important for
electron system. This electron donation will stabilize the meta derivatives, where it is only possible to
the non-ionized acid via electron delocalization, but donate electrons towards the ring carbons, which
would destabilize the conjugate base by creating a are, of course, not as electronegative as oxygen.
double charge in the carboxylate system. The net In fact, meta substitution is the least complicated,
result is lower acidity. in that groups placed there exert their influence
almost entirely through inductive effects. It should
This electron-donating effect from lone pair elec- be noted that, where we have opposing resonance
trons is simply a resonance effect, but is often termed and inductive effects, the resonance effect is normally
a mesomeric effect. A mesomer is another term for a of much greater magnitude than the inductive effect,

ELECTRONIC AND STRUCTURAL FEATURES THAT INFLUENCE ACIDITY 133

and its contribution predominates (but see below for CO2H O OH
chlorine).
>
The relatively high acidity of ortho-hydroxyben-
zoic acid (salicylic acid), compared with the Cl Cl
other derivatives just considered, is ascribed to
intramolecular hydrogen bonding, which is not strong weak
possible in the other compounds, even with ortho- inductive effect resonance effect
methoxybenzoic acid.

OO HO O O OH CO2H
H H >
O O

favourable H-bonding H-bonding in HO OH
stabilizes anion non-ionized acid
strong weak inductive
Hydrogen bonding involves a favourable six- resonance effect effect
membered ring and helps to stabilize the conjugate
base. Although some hydrogen bonding occurs in the Resonance can also influence the acidity of
non-ionized acid, the effect is much stronger in the hydroxyl groups, as seen in phenols. Cyclohexanol
carboxylate anion. has pKa 16, comparable to that of ethanol. On the
other hand, phenol has pKa 10, making it consid-
It should be noted that the electron-donating erably more acidic than a simple alcohol, though
resonance effects just considered are the result of lone less so than a carboxylic acid. This increased acid-
pair electrons feeding in to the π electron system. ity is explained in terms of delocalization of the
Potentially, any substituent with a lone pair might do negative charge into the aromatic ring system, with
the same, yet we did not invoke such a mechanism resonance structures allowing ring carbons ortho and
with chlorine substituents above. As the size of the para to the original phenol group to become electron
atom increases, lone pair electrons will be located in rich. Although the aromatic ring acts as an accep-
orbitals of higher level, e.g. 3p rather than 2p as in tor of electrons, and may be termed an electron sink,
carbon. Consequently, the ability to overlap the lone charge is dispersed towards carbon atoms, which is
pair orbital with the π electron system of the aromatic going to be less favourable than if it can be dis-
ring will diminish, a simple consequence of how persed towards more electronegative atoms such as
far from the atom the electrons are mostly located. oxygen.
Chlorine thus produces a low resonance effect but a
high inductive effect, and the latter predominates.

OH O OH O O OO

cyclohexanol phenol phenoxide charge delocalized towards
pKa 16 pKa 10 conjugate base ortho and para carbons

A good illustration of this concept is seen in a O O
series of nitrophenols. The nitro group itself has N N
to be drawn with charge separation to accommodate
the electrons and our rules of bonding. However, O O
resonance structures suggest that there is electron nitro group
delocalization within the nitro group.

134 ACIDS AND BASES

With substituted phenols, there can be similar delocalization of the negative charge of the phenoxide
delocalization of charge into the aromatic ring as conjugate base if it is situated in the ortho or para
with phenol, but substituents will introduce their positions. This increases acidity relative to phenol,
own effects, be it inductive or resonance related. and both compounds have essentially the same pKa
It can be seen that the nitro group allows further of 7.2.

OH OH OH OH OO OO
NO2 N N
phenol O O
pKa 10
o-nitrophenol NO2 NO2 resonance effects
pKa 7.2 p-nitrophenol O stabilize anions O
m-nitrophenol
pKa 8.4 pKa 7.2

OH OH N N
NO2 O2N NO2 OO OO

NO2 NO2 OO

2,4-dinitrophenol 2,4,6-trinitrophenol NO2 NO2
pKa 4.1 (picric acid)
pKa 0.4

inductive effect helps
to stabilize anion

The effect is magnified considerably if there are be quite pronounced. A summary list of resonance
nitro groups both ortho and para, so that the pKa effects emanating from various groups is shown in
for 2,4-dinitrophenol is 4.1. A third nitro group, as in Table 4.8. We should also point out that these very
2,4,6-trinitrophenol, confers even more acidity, and same principles will be used to rationalize aromatic
this compound has pKa 0.4, making it a strong acid.
This is reflected in its common name, picric acid. Table 4.8 Resonance effects from functional groups

Note that m-nitrophenol has pKa 8.4, and is a lot Electron-donating Electron-withdrawing
less acidic than o-nitrophenol or p-nitrophenol. We groups groups
can draw no additional resonance structures here, and
the nitro group cannot participate in further electron ––F ––C≡N
delocalization. The increased acidity compared with N
phenol can be ascribed to stabilization of resonance O
structures with the charge on a ring carbon through ––Cl C
the nitro group’s inductive effect.
––Br ––SR
From the above, it should not be difficult to
rationalize the effects of other types of substituent ––I ––SH ––SO2––
on the acidity of phenols. Thus electron-donating ––O ––CH3 ––NO2
groups, e.g. alkyl, reduce acidity, and electron- ––OR
withdrawing groups, e.g. halogens, increase acidity.
With strongly electron-withdrawing groups, such as ––OH
cyano and nitro, the acid-strengthening properties can
––OCOR

BASICITY 135

substitution reactions in Chapter 8, and this is why • a weak base has a large Ka and thus a small pKa,
we have purposely discussed the acidity of aromatic i.e. B is favoured over BH+.
derivatives in some detail.

Or, put another way:

4.4 Basicity • the larger the value of pKa, the stronger is the
base;
We have already defined a base as a substance that
will accept a proton by donating a pair of electrons. • the smaller the value of pKa, the weaker is the
Just as we have used pKa to measure the strength of base.
an acid, we need a system to measure the strength of
a base. Accordingly, a basicity scale based on pKb The relationship between pKa and pKb can be
was developed in a similar way to pKa. deduced as follows:

For the ionization of the base B in water [HO−][BH+]
[B]
K Kb =

B + H2O BH + HO [B][H3O+]
[BH+]
the equilibrium constant K is given by the formula Ka =

[HO−][BH+] Ka × Kb = [B][H3O+] × [HO−][BH+]
K= [BH+] [B]

[B][H2O] = [H3O+][HO−]

and since the concentration of water will be essen- Thus, Ka × Kb reduces to the ionization constant for
tially constant, the equilibrium constant Kb and the water Kw.
logarithmic pKb may be defined as

Kb = [HO−][BH+] K

[B] H2O + H2O H3O + HO

with base acid
pKb = − log10 Kb accepts donates
proton proton

This system has been almost completely dropped in In this reaction, one molecule of water is acting as
favour of using pKa throughout the acidity–basicity a base and accepts a proton from a second water
scale. To measure the strength of a base, we use molecule. This second water molecule, therefore,
the pKa of its conjugate acid, i.e. we consider the is acting as an acid and donates a proton. The
equilibrium equilibrium constant K for this reaction is given by
the formula
K
K = [H3O+][HO−]
BH + H2O H3O + B [H2O][H2O]

conjugate and because the concentration of water is essentially
constant in aqueous solution, the new equilibrium
acid constant Kw is defined as

for which [B][H3O+] Kw = [HO−][H3O+]
[BH+]
Ka =

It follows that

• a strong base has a small Ka and thus a large pKa, For every hydronium ion produced, a hydroxide anion
i.e. BH+ is favoured over B; must also be formed, so that the concentrations of